Time for some shenanigans @gsuberland
=> View attached media | View attached media | View attached media
=> More informations about this toot | More toots from jpm@aus.social
Obviously designed to be dropped, big rubber end caps. They just pull straight off, revealing 4 screws on each end holding the face plates on
=> More informations about this toot | More toots from jpm@aus.social
Crack it open!
4 RF cages covering the analogue front ends on the left, connecting to a 4-channel 1Gsps ADC in the middle, connecting to a mystery FPGA in the middle, connecting to a Cypress FX2A USB device controller on the middle lower edge. Some blank footprints where other device variants are populated. Does this look like the guts of yours @gsuberland ?
=> More informations about this toot | More toots from jpm@aus.social
@gsuberland there appears to only be one flash chip on the board, and it’s connected to the FX2A USB device controller. The JTAG header nearby is connected to the FPGA. I’ll be attacking this part of the board with my #GlasgowInterfaceExplorer soon, to get a dump of that flash chip and figure out what the FPGA is.
=> More informations about this toot | More toots from jpm@aus.social
Because what the FPGA is, is a mystery. If you search for “Lemontree 2090” the only results you get are from folks on the EEVBlog forums who are tearing down Hantek oscilloscopes.
There is some speculation that it’s a Gowin FPGA that’s either been white-labeled or customized. That’s what the digging on the JTAG connector will be for…
=> More informations about this toot | More toots from jpm@aus.social
ADC is an Analog Devices HMCAD1511 - 1 to 4-channel 1GSPS (1 channel, 250MSPS 4-channel) 8-bit with LVDS output. It’s a single-chip ADC designed for oscilloscopes
=> More informations about this toot | More toots from jpm@aus.social
Analogue frontend mystery chips: TI35 12ST, and 27E1 6552
Guessing the second is an op-amp, looks like there’s an alternate footprint there too
=> View attached media | View attached media
=> More informations about this toot | More toots from jpm@aus.social
Once again the #GlasgowInterfaceExplorer impresses. Zero issues dumping the MX25L8006E SPI flash chip using the memory-25x applet once I had it all connected and powered correctly, but no idea what the contents actually is. My guess is firmware for the FX2A, so 8051 assembly? Let me know if you want to poke at it and I’ll drop it somewhere
=> More informations about this toot | More toots from jpm@aus.social
@jpm Did the 8-pin in circuit probe work well for you? I did a lot of experiments like this, with an STM32 based DIY Flashrom device and the clamp only worked once every twenty tries.
It wasn't even a problem about the chip still being in circuit (it worked with a soldered in "programming connector), but more with the improper contacts made by the clamp.
Flashrom would recognize the chip, start flashing, and then have checksum errors when reading back to verify.
Is the glasgow more reliable?
=> More informations about this toot | More toots from bastelwombat@chaos.social
@bastelwombat they’re a bit annoying to use, and can’t be used if there’s other components in the way. They are designed to sit the tip flush on the PCB, pins being held in slots. This is a very cheap AliExpress special so is a bit more annoying to use, many folks recommend Pomona clips, eg https://www.digikey.com/en/products/detail/pomona-electronics/5250/745102
The Glasgow has been very reliable reading and writing common flash modules for me. The default communication speed is quite slow, which is ideal for using probe clips as high-speed signal integrity isn’t usually achievable.
=> More informations about this toot | More toots from jpm@aus.social
text/geminiThis content has been proxied by September (3851b).