TR-505 Bending & rom Hacking

8-Banks TR-505 MOD "TREX-2" and Flash ROM Programmer kit are still 5 pieces and 3 pieces available. I also ship overseas with Paypal payment, so please contact me by email if you want.
>> gadget_factory@bf.wakwak.com
Documents are here: https://drive.google.com/drive/folde...K0GiElz7rhJHmn

aaa.jpg ccc.jpg bbb.png
I want to sell my inventory early.

[playthatbeat]

Been a while!

So there are products for the 505 now.

There's a rom switching board: https://hkadesign.org.uk/tr505expansion.html

The same guy also does a board for the tr626 - this will allow banks of sounds to be switched in & out.

MEANWHILE, IN JAPAN: https://page.auctions.yahoo.co.jp/jp/auction/k490278098

Same story but a kit.

Same guys also make a 'blower' kit - so You can change the contents of the ROM with USB!

505blower1.jpg 505blower2.jpg 505blower3.jpg

Hard to say if these kits are available anywhere outside Japan. Vendor is called "gadget_factory_y" on Yahoo Japan https://auctions.yahoo.co.jp/seller/gadget_factory_y




Docs for kit, and files for his Rom are here: https://drive.google.com/drive/folde...K0GiElz7rhJHmn

If anyone manages to get any of this going, please share experience here!

[bartbral]

I'm sure there is scope to mod these circuits

http://www.mjcaselden.com/hacks-and-...-modification/

I have tried this... but it doesn't cut the mustard for me...
As he explains in the article:

As always, there is a downside. Excessive capacitance will dull out the transients, resulting in a loss of a crisp audible attack on some envelopes.

So I'm not going this route.
I have breadboarded some other simple AR-envelopes to go round this problem... But not pleased with the results yet. I'll report here if might find the holy grail.

even in the rolands where although the same styles of circuits are there, they aren't laid out in lovely obvious lanes.

The envelope is very basic and a doable mod I still think... But still can't put my head round the "filter" they have used. I might try to f#(k about with the resistors after the multiplexer in their so called S/H in the near future. It looks from the schematic that they are doing some kind of lowpass filter (different ohms for different sounds). But first I'll try to get an AR EG going... Or just solder back the envelope caps. Before I'll loose them on my too small and thrashy workspace.

[bartbral]

I have noticed this on the Oberhiem DX also - some roms intended for other machines, like the Linndrum roms will whine when the machine first boots, once you fire the voice the whine goes away for the remainder of the session.

This is where I found the idea... http://www.electrongate.com/dmxfiles...ade/index.html It has the "Auto Zero first two bytes option". I first discarded it and thought it was something I did wrong... Or that it was some clocksignal leaking through, so it was on the bottom of my "try this out list"... Then after trying every other possible cause I tried it out... And voila it was gone.

I first noticed it was gone after triggering the first note as well... But I wanted to get rid of it because I was afraid it would be there when I play live... And i'm using a 040 flashrom and didn't want it to be back after switching kits.

Interesting observation - this must be a 'zero crossing' style fix perhaps? So not only to avoid a click at the start of a sound, but to reset the state of the DAC or something? would save some cycles and some work while playing maybe?

I was thinking it might have something to do with that... But it's just a resistor ladder DAC so don't think that is the cause.

My first idea was that the multiplexer will scan the DAC's output... And it will find data on the first byte and will output that. Scanning all the outputs it will output a sequence of bytes and thus producing a tone.
Or the sample and hold Caps after the multiplexer might be charged with some random voltage, that will be in a known state after the first note triggering.

Whatever it is... I'm almost certain it has something to do with clock-C... because that is 40µs = +/-25kHz, and that is where my whining is in the frequency-spectrum.

LOL.. man, this has been the case for YEARS and i never bothered to ask WHY since it just goes away once you play the voice once.

I just changed my script to put some hex80's in front of my samples... So it won't be worrying me any more.

[playthatbeat]

Hi there... I don't know if someone is still following this post, but I have a question.
Trying to decode the samples, I found that they have a strange volume envelope, like a hard compressed sound not fading out like is heard on the real output of the machine itself.

So, there is someone who knows how sounds are faded in the TR-505, and if there's a way to have the same fade applyied on the samples from the EPROM?

All the 8bit romplers had hardware envelopes and filtering in the voice chains. Not sure how it's done on the Rolands (never really looked), but the more discrete/larger sized earlier machines (DX, LM2 etc.) you can clearly see the circuits laid out in a visually sensible way, and you can follow the audio all the way from the output of the DAC to the jacks for the individual outputs, through some well known LPF and VCA designs. Moving chips around from place to place inside the machine would result in different tonal quality and decays and so on for the same chips/sounds. I'm sure there is scope to mod these circuits, even in the rolands where although the same styles of circuits are there, they aren't laid out in lovely obvious lanes.

[playthatbeat]

Interesting observation - this must be a 'zero crossing' style fix perhaps? So not only to avoid a click at the start of a sound, but to reset the state of the DAC or something? would save some cycles and some work while playing maybe?

I have noticed this on the Oberhiem DX also - some roms intended for other machines, like the Linndrum roms will whine when the machine first boots, once you fire the voice the whine goes away for the remainder of the session.

LOL.. man, this has been the case for YEARS and i never bothered to ask WHY since it just goes away once you play the voice once.

Hmmmmmmm (deep thought face)

[bartbral]

And I wanted to add a hint for the people (like me) who noticed a high pitched whining sound in their own sample packs.

If you look at the original samples you'll notice that some samples have a bit of silence added in front of the samples. I first thought the Roland people just didn't didn't cut the samples really carefully... But that isn't the case!

Put the same silence (hex 80) in the beginning of your own samples!!!
This will solve the high pitched whine!!!

This took me a while to solve...
So do not try to re-invent the wheel...

[bartbral]

Changed Your quota x 10..

CHEERS NICE ONE!




edit: Oooops!!! a lot of nonsense deleted (The thumbnail image is not correct... It ain't a RC low-pass filter... i think)

[playthatbeat]

Also:

all the chip/pcm drum machines i have played about with have some form of filtering after the DAC.. even something like the TR909 has a simple capacitor filter to take the edge off the samples used for the cymbals, and one of the service bulletins for it mentions changing a cap value to remove some low-end. Comparing the raw PCM to a sample from the main outs is the only way to go really, then look at them as slices in a spectrum analyser, raw vs. output. You can imagine what the filtering is going to do - it's likely a simple envelope in hardware, fixed attack and decay, used to smooth out a grainy sample with a LPF (or HPF in some cymbals)

Something like the oberhiem DX/DMX looks like it has the same hardware used again and again per channel, just with different samples on each, but if you move a chip from (lets say) the snare to the tomtom slot, you can clearly hear it sounds different. Same story when using Linndrum chips in something other then a Linndrum - they don't sound the same at all, but can get very close if you use them in like-for-like slots.

[playthatbeat]

I hope this one is correct. Because I am near to my max attachment KB's

Changed Your quota x 10..

[bartbral]

So, there is someone who knows how sounds are faded in the TR-505

tr505_env_section.jpg

Edit: I forgot the most important in the picture:
The Capacitors:
C20 (SNARE Decay (0.022))
C21 (KICK Decay (0.022))
C22 (CLAP & RIMSHOT Decay (0.022))
C23 (HIHAT Decay (0.1))
C24 (CRASH & RIDE Decay (0.1))
C25 (HI & LOW COWBELL Decay (0.022))
C26 (HI, MID & LOW TOMS + TIMBAL Decay (0.047))
C27 (HI & LOW CONGA Decay (0.022))

and if there's a way to have the same fade applyied on the samples from the EPROM?

Sample original tr-505 sounds from the output.
Compare the applied envelope to that of the ROM .BIN raw sample.

or...

Put a test tone into all of your romsamples. (sinewave or noise).
then sample your output and compare that with the input samples.
Then in software make an envelope that will boost the attack so that it will match the output envelope.

Hit and miss... Experiment...

[ragazzobcn]

Hi there... I don't know if someone is still following this post, but I have a question.
Trying to decode the samples, I found that they have a strange volume envelope, like a hard compressed sound not fading out like is heard on the real output of the machine itself.

So, there is someone who knows how sounds are faded in the TR-505, and if there's a way to have the same fade applyied on the samples from the EPROM?

[bartbral]

taking a closer look... some samples where still interleaved... and some samples needed there bytes swapped...

I hope this one is correct. Because I am near to my max attachment KB's

[bartbral]

Screen shot 2015-02-06 at 12.15.18 AM.jpgUpdate...

Here's the Raw Rom dump... (with the heavy metal and all hats)

And the unscrambled TR-505 Kit...


To be continued........

[bartbral]

I don't know if your reader supports the:
Toshiba TC571001D,
or
Hitachi HN27c301
or
NEC PD27C1000
SGS-Thompson (ST) M27C1000
but NOT the Macronix (MXIC) MX27c1000

Those are all pin by pin compatible EPROMs. They have 32 pins... But all Address- and Data-lines are exactly the same. Most (standard / JEDEC) compatible EPROMs have /OE and A16 swapped from the Toshiba TC531000AP mask-rom... It would be a possible cause for the missing cymbals...

==>> PIN COMPATIBLE EPROM !!! <<==

(1M) 128k x 8-bit EPROM                 

Vpp  1 +-v-+ 32  Vcc
OE/  2 |   | 31  PGM/
A15  3 |   | 30  nc
A12  4 |   | 29  A14
A7   5 |   | 28  A13
A6   6 |   | 27  A8
A5   7 |   | 26  A9
A4   8 |   | 25  A11
A3   9 |   | 24  A16
A2  10 |   | 23  A10
A1  11 |   | 22  CE/
A0  12 |   | 21  D7
D0  13 |   | 20  D6
D1  14 |   | 19  D5
D2  15 |   | 18  D4
GND 16 +---+ 17  D3

I just took a quick peek...and it seems my Eprom-reader to be... Doesn't support it... But well it was only €50,- so I shouldn't complain... So I will have to make an adapter Swapping those pins ( A16 <==> OE/ )... and the Vcc-pin

[edit: found another pin compatible EPROM]
[edit: and another...]