bookishboy
New Member
Hi, I posted this same question over at the electro-music forums about a week ago, but the thread hasn't generated much useful information so far. I'm hoping that someone here may be able to answer my questions or walk me through how to properly find the answers myself.
I suspect that the answer will come from:
-Someone who's familiar with the workings of *very* cheap electronic music devices, mostly which would be sold in less affluent countries, or
-Someone who has worked in the toy industry, particularly toys that talk or play sounds/music from an electronic chip, or
-Someone who's worked with electronic sound formats for a long time and is familiar with formats no longer used on commercial music players (not .wav, not .mp3, not .wma, etc)
I own a Buddha Machine ambient music loop player which mostly stopped working, apparently because it was an early production model with lower assembly standards. I'm going to have it circuit-bent and modded, which should hopefully get it working again and make it more interesting than it already is. As part of the mod, I'm trying to see if it's possible to have the unit's memory swapped out with something like a SD card reader. If I can, and also find what format the existing music is encoded in, it could be possible to put different loops into there. This would open up some very interesting possibilities, including potentially a project where people could create their own players.
I have very little knowledge of electronics. Everything that I've been able to figure out about the workings of the device and its ROM chip have been though reading articles about the device, taking it apart, googling part numbers, and making educated guesses.
The device is described here:
FM3 BUDDHA MACHINE
..and in an interview here:
Rare Frequency › The Ghost in the Buddha Machine: An Interview with Christiaan Virant of FM3
The exploded graphic on the front page is actually of a newer version of the machine..... still close though. A similar graphic of my box's internals can be found here:
http://www.fm3buddhamachine.com/downloads/buddhaexploded_300dpi.jpg
In a nutshell, the thing contains 9 sound loops on a burn-once ROM chip. You can turn it on, turn up or down the volume, and activate a dual-position switch which advances the device to the next loop. When turned on, the device starts playing loop#1 and continues to do so endlessly until it's switched off, advanced to the next loop, or loses power. Its design follows very closely on that of buddhist "chant boxes", "chanting machines" or "buddhist jukeboxes", several of which can be found here:
**broken link removed**
The overall design of the device is very simple. It's powered by 2AA batteries, or alternately an external 3v dc source (a plug/wall-wart inverter). Sound is produced from a single integral speaker, or alternately split and delivered to a set of headphones. Power on/off and volume is from the same dial. Finally, there's a dual-position swich which advances to the next track in line each time it's flicked back and forth.
Aside from the inputs/outputs and interface components, I found only 2 chips on the board. I was able to identify one as a SGS-Thompson TDA2822M, whose data sheet can be found here:
http://www.st.com/stonline/products/literature/ds/1464/tda2822m.pdf
The other chip is an unmarked square which juts up perpendicular from the main pcb. (in the exploded diagram of the device, it can be seen to the rear of the seated buddha, just behind 2 resistors.) It has 5 traces etched into it which connect(soldered) to traces on the main pcb. The center of the chip is covered with a shallow black blob, presumably a protective coating. I assumed (later confirmed) that this is the ROM chip which contains the sound files.
Deductions:
-The TDA2822M chip's datasheet lists its function as stereo amplification but not digital signal processing. Therefore I think it's likely that whatever format the music file(s) are in on the ROM chip, they don't require interpretation, they simply play, directly as recorded sound.
-The dual-position switch on the unit's side advances from one track to the next. The tracks are of wildly varying length, from a few seconds to a few minutes. This suggests to me that there are discrete files and perhaps a file structure to be found on the ROM chip.
Potentially of use:
-One of the system's two conceptual designers, when contacted, replied that he wasn't sure of all the technical specifications including file formats. He did confirm that the square chip was the ROM chip where the music file(s) reside, and that "the music encoding is some non-standard format, used mostly in the chip fab industry. before our music is burned onto a large memory master, its encoded in this format".
-There are 2 slots cut into the main pcb, one on each side of the ROM chip, with traces leading to each. Nothing occupies these slots in the assembled machine, which leads me to suspect that they're either for auxiliary memory (more ROM chips) or they're connection points between the pcb and a ROM-burning apparatus. If the second were true, the assembled board would be hooked up to the apparatus and the integrated ROM chip burned in situ, instead of being burned first and then soldered onto the pcb.
-"When we made the loops longer and tested them on a higher-capacity chip, it sounded awful — something to do with the clock speed of the chip and how it interacted with the PC board. So we went back to the lower-capacity chip, which forced us to squeeze everything into 300 seconds of music." (referring to the design of the still-similar 2.0 box)
-"Those may even be WAV files. Those were the original files given to the factory, so you can hear the difference in what they sounded like when we gave 'em to the factory and what they ended up like on the little 8-bit audio chip inside the box." (referring to the design of the 1.0 box)
Goals, if feasible:
-determine what format the data is in on the ROM chip.
-replace the chip with a flash memory reader such as a SD card reader. Alternately find a way to supplement the device's existing memory with a SD card reader (this only possible if the two extra "slots" on the pcb were designed for extra memory, instead of for burning data to the single ROM.
-Find a Windows-based software encoder which allows me to encode other .wav or .mp3 sound samples into the 8-bit format recognized by this player.
-Potentially, determine the clock speed of the ROM chip and determine how a consumer-grade flash memory format would compare, both on a technical level and in sound quality.
Sorry for the overlong first post. I'm hoping that somewhere on this board there is someone who's familiar enough with audio circuitry design that they'll be able to steer me in the right direction. Any takers?
Thanks for any help. Smile
I suspect that the answer will come from:
-Someone who's familiar with the workings of *very* cheap electronic music devices, mostly which would be sold in less affluent countries, or
-Someone who has worked in the toy industry, particularly toys that talk or play sounds/music from an electronic chip, or
-Someone who's worked with electronic sound formats for a long time and is familiar with formats no longer used on commercial music players (not .wav, not .mp3, not .wma, etc)
I own a Buddha Machine ambient music loop player which mostly stopped working, apparently because it was an early production model with lower assembly standards. I'm going to have it circuit-bent and modded, which should hopefully get it working again and make it more interesting than it already is. As part of the mod, I'm trying to see if it's possible to have the unit's memory swapped out with something like a SD card reader. If I can, and also find what format the existing music is encoded in, it could be possible to put different loops into there. This would open up some very interesting possibilities, including potentially a project where people could create their own players.
I have very little knowledge of electronics. Everything that I've been able to figure out about the workings of the device and its ROM chip have been though reading articles about the device, taking it apart, googling part numbers, and making educated guesses.
The device is described here:
FM3 BUDDHA MACHINE
..and in an interview here:
Rare Frequency › The Ghost in the Buddha Machine: An Interview with Christiaan Virant of FM3
The exploded graphic on the front page is actually of a newer version of the machine..... still close though. A similar graphic of my box's internals can be found here:
http://www.fm3buddhamachine.com/downloads/buddhaexploded_300dpi.jpg
In a nutshell, the thing contains 9 sound loops on a burn-once ROM chip. You can turn it on, turn up or down the volume, and activate a dual-position switch which advances the device to the next loop. When turned on, the device starts playing loop#1 and continues to do so endlessly until it's switched off, advanced to the next loop, or loses power. Its design follows very closely on that of buddhist "chant boxes", "chanting machines" or "buddhist jukeboxes", several of which can be found here:
**broken link removed**
The overall design of the device is very simple. It's powered by 2AA batteries, or alternately an external 3v dc source (a plug/wall-wart inverter). Sound is produced from a single integral speaker, or alternately split and delivered to a set of headphones. Power on/off and volume is from the same dial. Finally, there's a dual-position swich which advances to the next track in line each time it's flicked back and forth.
Aside from the inputs/outputs and interface components, I found only 2 chips on the board. I was able to identify one as a SGS-Thompson TDA2822M, whose data sheet can be found here:
http://www.st.com/stonline/products/literature/ds/1464/tda2822m.pdf
The other chip is an unmarked square which juts up perpendicular from the main pcb. (in the exploded diagram of the device, it can be seen to the rear of the seated buddha, just behind 2 resistors.) It has 5 traces etched into it which connect(soldered) to traces on the main pcb. The center of the chip is covered with a shallow black blob, presumably a protective coating. I assumed (later confirmed) that this is the ROM chip which contains the sound files.
Deductions:
-The TDA2822M chip's datasheet lists its function as stereo amplification but not digital signal processing. Therefore I think it's likely that whatever format the music file(s) are in on the ROM chip, they don't require interpretation, they simply play, directly as recorded sound.
-The dual-position switch on the unit's side advances from one track to the next. The tracks are of wildly varying length, from a few seconds to a few minutes. This suggests to me that there are discrete files and perhaps a file structure to be found on the ROM chip.
Potentially of use:
-One of the system's two conceptual designers, when contacted, replied that he wasn't sure of all the technical specifications including file formats. He did confirm that the square chip was the ROM chip where the music file(s) reside, and that "the music encoding is some non-standard format, used mostly in the chip fab industry. before our music is burned onto a large memory master, its encoded in this format".
-There are 2 slots cut into the main pcb, one on each side of the ROM chip, with traces leading to each. Nothing occupies these slots in the assembled machine, which leads me to suspect that they're either for auxiliary memory (more ROM chips) or they're connection points between the pcb and a ROM-burning apparatus. If the second were true, the assembled board would be hooked up to the apparatus and the integrated ROM chip burned in situ, instead of being burned first and then soldered onto the pcb.
-"When we made the loops longer and tested them on a higher-capacity chip, it sounded awful — something to do with the clock speed of the chip and how it interacted with the PC board. So we went back to the lower-capacity chip, which forced us to squeeze everything into 300 seconds of music." (referring to the design of the still-similar 2.0 box)
-"Those may even be WAV files. Those were the original files given to the factory, so you can hear the difference in what they sounded like when we gave 'em to the factory and what they ended up like on the little 8-bit audio chip inside the box." (referring to the design of the 1.0 box)
Goals, if feasible:
-determine what format the data is in on the ROM chip.
-replace the chip with a flash memory reader such as a SD card reader. Alternately find a way to supplement the device's existing memory with a SD card reader (this only possible if the two extra "slots" on the pcb were designed for extra memory, instead of for burning data to the single ROM.
-Find a Windows-based software encoder which allows me to encode other .wav or .mp3 sound samples into the 8-bit format recognized by this player.
-Potentially, determine the clock speed of the ROM chip and determine how a consumer-grade flash memory format would compare, both on a technical level and in sound quality.
Sorry for the overlong first post. I'm hoping that somewhere on this board there is someone who's familiar enough with audio circuitry design that they'll be able to steer me in the right direction. Any takers?
Thanks for any help. Smile