Circuitbenders Forum
Circuitbenders Forum => Drum machines => Circuitbending discussion => Roland & Boss => Topic started by: sensor on November 17, 2008, 12:52:39 PM
-
my 626 makes only digital noise now. it sounds cool but it is really not what i want. i think it has something to do with the dac.
is there anybody out there who could give me a copy of the schematics, i just haven't found them in the net. or does anybody knows the problem and how to fix it?
thanks,
matthias
-
I've got a broken 626 if you need spares. The DAC is the resistor pack connected to the data lines of the big EPROM with the samples in it. It's just an R/2R ladder, as far as I can see. IIRC eight pins go to the data lines of the EPROM, one goes to ground and one is the signal out that is passed through the nearby opamp and off to the multiplexer that splits out the separate outputs.
Unfortunately the control chips are very fragile, and if they're sending garbage to the EPROM address lines, you'll get weird digital noises. If the samples sound basically okay but are distorted, it might be just the EPROM or a poor connection to the resistor pack.
-
thanks a lot for your answer.
well, it makes noise. the rhytms are ok, but there really nothing sounds like a bassdrum or a hat. so it is very hard to say what is wrong. spares were quite cool.
matthias
-
well it's the eprom i think. it was really cool if you could send me your eprom, dac and the doa. perhaps you could contact me via myspace.com/burningw.
matthias
-
Have you got access to an EPROM programmer? That way you can just burn a new EPROM, or even blow it into flash memory.
-
yes i 've got access to an eprom burner. but where to get the data?
matthias
-
From here: http://www.gjcp.net/~gordonjcp/tr626.bin.gz
Don't worry if you haven't got a ROM big enough. I used to test custom sample sets with a 28F1000 1M Flash chip, instead of the 2M that the 626 needs. All it means is that some of the samples appear twice (you've only got one bass drum, not two).
-
well, i'll try to burn a new eprom and see what happens.
how do i get custom sample sets into the eprom? and how to switch between different eproms? is there a chip enable pin or something?
matthias
-
The ROM contains the samples in a weird interleaved pattern, which you need to create with a utility I wrote (and can't seem to find right now!)
If you wanted to use more than one ROM you'd stack the two chips one on top of the other, with all the pins except /OE or /CE connected together. Then you'd use a pull-up resistor on /CE (or /OE) to make sure the chip stayed "off", and use a toggle switch to ground the appropriate pin to select the chip. This would work for *any* drum machine ROM.
If you found a drum machine (or indeed anything else with a ROM) where you could fit a ROM twice as big into the same socket (eg. going from 27C256 to 27C512) then you could blow two images into the ROM one after the other, and use the same pull-up resistor and switch arrangement on the highest address pin, to switch which "half" of the ROM was in use.
-
it was pretty cool if you find the programm for making custom sound sets. i'm bad in programming.
donation was no problem.
matthias
-
do yoy have some data sheet of the roland rom chip or some pin function list?
matthias
-
It's just a JEDEC ROM, something like 2Mbit. Look for something like a 27C2001.
-
So hold on, am I understanding this thread correctly? Are you folks saying that it's possible to write your own samples onto a new chip and install the new chip in the tr-626 drum machine?
Wow, I didn't know that was possible!
I have a tr-505 which I think is similar to the 626, do you think it'd be possible to replace the drums in it with my own custom samples?
It's cirtainly beyond my abilities right now, but if anyone could point me in the direction of a simple tutorial then I'd love to give this a try. Am I correct in thinking that it's possible to add new chips/drums to the machine without destroying the original drums? .....Just using a switch to select between chips?
I suppose that the circuit-bends would function in the same way with new chips, right?
That is very cool indeed. :o
-
Yes. It's not *that* simple because the samples are scattered all over the place in the 626 ROM. I've actually got a 505 thatI intended to decode the ROM in and write a ROM-mangler for. The chip used in the 505 is slightly different, and would need an adaptor socket made up to suit the larger chips that you could use for extra sample banks.
As with all ROM samples, you could use a pull-up resistor on the /CE line and some sort of gating circuit to select which chip you used.
-
You sound like a really smart guy Gordonjcp, how did you figure this stuff out (if I'm not been cheeky asking)?
Could you reccomend any good 'entry point' books to read on the subject of eprom writing/decoding/hacking/etc. ?
I really fancy giving this a try with my 505, but right now it's far out of my understanding. Gotta get some education on the subject first! :D
-
Experiment... I can't remember where I got the spec for the 505 ROM, but most ROMs follow a fairly standard "pattern" of pin connections. Actually decoding the ROM was partly guesswork, partly experimentation and partly measuring the frequencies of the pulses on the TR626 ROM address lines.
Lots of beer helps, too. I recommend a fairly light pale ale to start with, perhaps something like Badger Dorset Ale, or maybe Harviestoun Bitter & Twisted, or even Houston Killellan (not really an IPA but still worth a go). As you get deeper into the mysteries of the ROM you might try something like Badger Tanglefoot or maybe Kelburn Red Smiddy (quite light, you'll need a lot of this) or of course Orkney Dark Island. Obviously if these beers are unavailable then you can substitute some locally-available real ale.
Really it comes down to this: have a pencil and paper, and know how to use it; have some real ale, and know how to use it; and have an oscilloscope or signal tracer, and know how to use it.
Hope this helps. Good luck!