Furrtek
@furrtek.bsky.social
Virtual coyote obsessed with retro electronics reverse-engineering, theme park tech, music, and giant robots.
Interested in the demoscene, SDR, industrial urbex, pointy animals, graffiti, things on rails, North Korea 💚. Talks 🇫🇷/🇺🇸
Interested in the demoscene, SDR, industrial urbex, pointy animals, graffiti, things on rails, North Korea 💚. Talks 🇫🇷/🇺🇸
The gaps are normal, I was talking about these
November 15, 2025 at 6:08 AM
The gaps are normal, I was talking about these
No, the glitchy values wouldn't fit better anywhere else. Plus the busses used for readout are also used for testing the multipliers and those give out all the expected values, so I'm quite sure about the bit order.
November 15, 2025 at 12:18 AM
No, the glitchy values wouldn't fit better anywhere else. Plus the busses used for readout are also used for testing the multipliers and those give out all the expected values, so I'm quite sure about the bit order.
Dumped the internal ROMs, near perfect match with the optical readout.
Glitchy values in volume LUT confirmed, nice job Konami 👍
Glitchy values in volume LUT confirmed, nice job Konami 👍
November 14, 2025 at 3:56 AM
Dumped the internal ROMs, near perfect match with the optical readout.
Glitchy values in volume LUT confirmed, nice job Konami 👍
Glitchy values in volume LUT confirmed, nice job Konami 👍
What's wrong Konami ? You haven't even touched your top 5 bits.
This is before the final digital volume control so ideally it should be max amplitude. Feeling like I'm missing something.
This is before the final digital volume control so ideally it should be max amplitude. Feeling like I'm missing something.
November 11, 2025 at 8:08 AM
What's wrong Konami ? You haven't even touched your top 5 bits.
This is before the final digital volume control so ideally it should be max amplitude. Feeling like I'm missing something.
This is before the final digital volume control so ideally it should be max amplitude. Feeling like I'm missing something.
Pairs of registers defining the speed and amplitude of a triangle waveform generator, which is used as an offset from the reverb buffer's tail: Heavy suspicion of an unemulated flanger effect in the Konami 054539.
November 3, 2025 at 1:01 PM
Pairs of registers defining the speed and amplitude of a triangle waveform generator, which is used as an offset from the reverb buffer's tail: Heavy suspicion of an unemulated flanger effect in the Konami 054539.
"The 054539 is quite complicated, maybe making a data flow diagram will help me understand how things are supposed to work"
...sheeeeeeit
...sheeeeeeit
October 29, 2025 at 7:12 AM
"The 054539 is quite complicated, maybe making a data flow diagram will help me understand how things are supposed to work"
...sheeeeeeit
...sheeeeeeit
Hello ma'am
That's a nice RF remote you got there
That's a nice RF remote you got there
August 9, 2025 at 5:47 PM
Hello ma'am
That's a nice RF remote you got there
That's a nice RF remote you got there
Dragon ! He sneezed on me :(
1/2
1/2
August 9, 2025 at 5:47 PM
Dragon ! He sneezed on me :(
1/2
1/2
[CloneCD success.wav plays loudly]
August 1, 2025 at 11:38 PM
[CloneCD success.wav plays loudly]
I/Q of radio frame. Obviously PSK, but what's with the noise ? It occurs at the same points while the frame is repeated. Can something on the receiver side cause this ?
April 20, 2025 at 10:50 AM
I/Q of radio frame. Obviously PSK, but what's with the noise ? It occurs at the same points while the frame is repeated. Can something on the receiver side cause this ?
IR photoreceptor 5-in-1 value pack
March 30, 2025 at 6:51 PM
IR photoreceptor 5-in-1 value pack
Updated the RE'd Fujitsu AV gate array cell library.
It's now 73% complete.
github.com/furrtek/Sili...
It's now 73% complete.
github.com/furrtek/Sili...
March 24, 2025 at 8:26 AM
Updated the RE'd Fujitsu AV gate array cell library.
It's now 73% complete.
github.com/furrtek/Sili...
It's now 73% complete.
github.com/furrtek/Sili...
If you want to perform the fix, take the converter apart (4 screws), cut the traces following the red mark, and swap them with two short wires soldered pad to pad. 4/4
March 5, 2025 at 10:51 PM
If you want to perform the fix, take the converter apart (4 screws), cut the traces following the red mark, and swap them with two short wires soldered pad to pad. 4/4
The Basement Brothers have discovered an error that went undetected for years in my Fusion AES to MVS converter ! It affects indie/homebrew releases only.
Fixing it allows Super Bubble Pop, and NG:DT's Neo Xyx, Gunlord and Razion to work. The Neo SD Pro still isn't happy. 1/4
Fixing it allows Super Bubble Pop, and NG:DT's Neo Xyx, Gunlord and Razion to work. The Neo SD Pro still isn't happy. 1/4
March 5, 2025 at 10:51 PM
The Basement Brothers have discovered an error that went undetected for years in my Fusion AES to MVS converter ! It affects indie/homebrew releases only.
Fixing it allows Super Bubble Pop, and NG:DT's Neo Xyx, Gunlord and Razion to work. The Neo SD Pro still isn't happy. 1/4
Fixing it allows Super Bubble Pop, and NG:DT's Neo Xyx, Gunlord and Razion to work. The Neo SD Pro still isn't happy. 1/4
Completed the reverse-engineering of the Konami 053936 "3D" plane generator chip used by a bunch of GX games, which had several undocumented settings.
One thing less to do !
github.com/furrtek/Sili...
One thing less to do !
github.com/furrtek/Sili...
January 29, 2025 at 1:38 PM
Completed the reverse-engineering of the Konami 053936 "3D" plane generator chip used by a bunch of GX games, which had several undocumented settings.
One thing less to do !
github.com/furrtek/Sili...
One thing less to do !
github.com/furrtek/Sili...
Konami Open Golf Championship "voxel" terrain height data.
I'd work on the rendering chip if more games used it, and if I didn't swear 2025 was a finish-things-don't-start-anything-new year.
I'd work on the rendering chip if more games used it, and if I didn't swear 2025 was a finish-things-don't-start-anything-new year.
January 21, 2025 at 8:56 PM
Konami Open Golf Championship "voxel" terrain height data.
I'd work on the rendering chip if more games used it, and if I didn't swear 2025 was a finish-things-don't-start-anything-new year.
I'd work on the rendering chip if more games used it, and if I didn't swear 2025 was a finish-things-don't-start-anything-new year.
Selling Konami chips that I already have to recoup cost of a working sacrificial GX board: 2x 054539, 054338, 056879, 056766, 053246A, 054156, 055555 and 68EC020 CPU. All in perfect shape, ready to be used for repairs. 9€ each. 4€ tracked intl shipping for any number. PayPal only.
January 14, 2025 at 2:50 PM
Selling Konami chips that I already have to recoup cost of a working sacrificial GX board: 2x 054539, 054338, 056879, 056766, 053246A, 054156, 055555 and 68EC020 CPU. All in perfect shape, ready to be used for repairs. 9€ each. 4€ tracked intl shipping for any number. PayPal only.
More (late) die shots, this time from 🐝
HuC6280: PC-Engine CPU & PSG
siliconpr0n.org/map/hudson/h...
HuC6260: PC-Engine VCE
siliconpr0n.org/map/hudson/h...
HuC6201: Ten No Koe 2 (not a whole lot going on in there)
siliconpr0n.org/map/hudson/h...
HuC6280: PC-Engine CPU & PSG
siliconpr0n.org/map/hudson/h...
HuC6260: PC-Engine VCE
siliconpr0n.org/map/hudson/h...
HuC6201: Ten No Koe 2 (not a whole lot going on in there)
siliconpr0n.org/map/hudson/h...
December 3, 2024 at 11:49 PM
More (late) die shots, this time from 🐝
HuC6280: PC-Engine CPU & PSG
siliconpr0n.org/map/hudson/h...
HuC6260: PC-Engine VCE
siliconpr0n.org/map/hudson/h...
HuC6201: Ten No Koe 2 (not a whole lot going on in there)
siliconpr0n.org/map/hudson/h...
HuC6280: PC-Engine CPU & PSG
siliconpr0n.org/map/hudson/h...
HuC6260: PC-Engine VCE
siliconpr0n.org/map/hudson/h...
HuC6201: Ten No Koe 2 (not a whole lot going on in there)
siliconpr0n.org/map/hudson/h...
HackRF TRF37B73 TX amp transplant is a success. Hopefully this one won't be as fragile.
December 3, 2024 at 1:14 PM
HackRF TRF37B73 TX amp transplant is a success. Hopefully this one won't be as fragile.
*yotes your digital audio*
November 28, 2024 at 4:34 AM
*yotes your digital audio*
Why I dislike tracing Toshiba gate arrays episode 57:
This is an adder.
It adds an 8 bit number with a 24 bit one.
That's it. Nothing else.
This is an adder.
It adds an 8 bit number with a 24 bit one.
That's it. Nothing else.
November 16, 2024 at 10:52 AM
Why I dislike tracing Toshiba gate arrays episode 57:
This is an adder.
It adds an 8 bit number with a 24 bit one.
That's it. Nothing else.
This is an adder.
It adds an 8 bit number with a 24 bit one.
That's it. Nothing else.
Oh wtf, the datasheet even includes die pictures to help with identification !
November 10, 2024 at 1:43 AM
Oh wtf, the datasheet even includes die pictures to help with identification !
Found it ! It's a Generalplus GPBA01B, a specialized IO expander for parallel memories.
Power and clock pads match the datasheet's diagram.
Sucks when the die ID is different from the part number.
Power and clock pads match the datasheet's diagram.
Sucks when the die ID is different from the part number.
November 10, 2024 at 1:29 AM
Found it ! It's a Generalplus GPBA01B, a specialized IO expander for parallel memories.
Power and clock pads match the datasheet's diagram.
Sucks when the die ID is different from the part number.
Power and clock pads match the datasheet's diagram.
Sucks when the die ID is different from the part number.