Edit: Updated my previous post with the new flashrom windows binary.
Sometimes flashrom identifies several flash chip types, which need to be specified with the -c command:
flashrom
Calibrating delay loop… OK. Found PMC flash chip "Pm25LV010" (128 kB, SPI) on ch341a_spi. Found PMC flash chip "Pm25LV010A" (128 kB, SPI) on ch341a_spi. Multiple flash chip definitions match the detected chip(s): "Pm25LV010", "Pm25LV010A" Please specify which chip definition to use with the -c <chipname> option..
Read BIOS from a specified flash chip:
flashrom -c Pm25LV010 -r backup.bin
Calibrating delay loop… OK. Found PMC flash chip "Pm25LV010" (128 kB, SPI) on ch341a_spi. Reading flash… done.
Wow. Wootever, you are the best of the web. Someone who would go to all the work to build that, and then just throw it out there for anyone to use is a scholar AND a gentleman. Granted, that’s the whole FOSS community modus operandi, and kind of how the Bus Pirate came to be, but still…I’m seriously appreciative.
I guess I’m buying a CH341A now. I bought a TL866A instead because it supposedly supported ICSP, but then I got it home and found out that it only supports it on a handful of chips, and the uber-common 25SPI chips are not supported for ICSP and it won’t work. >:( I only didn’t buy it because it was late at night and I interpreted a warning about being unable to write to ESMT 25s as being unable to write all 25SPI chips. My bad. I’m working with Macronix so it should be just fine.
Is there any reason you can think of why it wouldn’t work with the MX25V4006E family? Datasheet says they run 2.5-3.6 volts, which puts the CH341A in spec. It’s on a Toshiba hard drive board.
@SilverPuppy Unfortunately just found out that the black pcb version supplies 5V to the chip due to an improper design (CH341A Programmer Schematic, CH341A Power Supply Fix). Might be possible that it would fry 3.3V chips that are less tolerant to 5V.
The green pcb version works correctly by supplying only 3.3V to all pins, but i would also recommend to order the 1.8V adapter as an secondary safety measure.
Green PCB:
Black PCB:
Pin Overview:
The voltage on CH341A VCC (5V or 3.3V) provides the same amount to CS, DO, CLK, DI
Edit: About the MX25V4006E, it seems that flashrom has no specific support, but the corresponding MX25L and MX25V versions seems interchangeable.
From flashchips.h:
#define MACRONIX_MX25L4005 0x2013 /* Same as MX25L4005A, MX25L4005C, MX25L4006E / #define MACRONIX_MX25L8005 0x2014 / Same as MX25V8005, MX25L8006E, MX25L8008E, FIXME: MX25L8073E (4k 0x20) */
Thank you very much Wootever, your work is very appreciated !
I’ve used the the black PCB version (with Flashrom, because with the Chinese flash program I had errors with 1.18, 1.29 and 1.30) on a Winbond 25Q128FV1Q with success, although the flash was somewhat slow, it worked great. I used it to disable the ME (Management Engine with) me_cleaner.
Since you said this programmer can fry some chips, I had ordered the green one you’ve linked.
edit: This is the correct placement of the chip in the ZIF socket:
I have placed it backwards at first, and the chip and the programmer ran very hot !
I had similar problems with the CH341A Programmer software, there are also various versions available with some of them cracked or bypassed and without any official download source. When looking for alternatives i stumbled across flashrom, which worked remarkably well. (I initially used a Ubuntu VM for flashrom, but that wasn’t very practical or portable.)
The slower speed might be a limitation of the WinUSB driver or even the CH341A itself.
The output pins of the CH341A (CS, DO, CLK, DI) are supplied with 5V on the black pcb version. It seems most of the desktop mainboard 3.3V flashchips will tolerate the overvoltage, but it might be dangerous to other components if used with a soic clip on notebooks for example.
Indeed, some of the earlier CH341A Programmer versions had a wrong pinout picture.
Actually, this image was taken from the V1.18, it’s the newer version (1.30) that have a wrong image. You can also see on the PCB it had the correct pinout engraved.
About the speed, the only version of the Chinese program that somewhat worked (v1.18) was even slower than flashrom, and it crashed a lot.
Small update: I added a simple timer to compare the flashing process time between different chips or programmers.
Fortunately, the programming speed for the CH341A is identical on Windows and Linux. Windows log:
flashrom -r backup.bin
Calibrating delay loop… OK. Found Winbond flash chip “W25Q64.W” (8192 kB, SPI) on ch341a_spi. Reading flash… done (in 1 minute and 4 seconds).
flashrom -E
Calibrating delay loop… OK. Found Winbond flash chip “W25Q64.W” (8192 kB, SPI) on ch341a_spi. Erasing and writing flash chip… Erase/write done (in 3 minutes and 13 seconds).
flashrom -w bios.bin
Calibrating delay loop… OK. Found Winbond flash chip “W25Q64.W” (8192 kB, SPI) on ch341a_spi. Reading old flash chip contents… done (in 1 minute and 4 seconds). Erasing and writing flash chip… Erase/write done (in 56 seconds). Verifying flash… VERIFIED (in 1 minute and 5 seconds). Process completed in 3 minutes and 5 seconds.
Linux log:
sudo flashrom -p ch341a_spi -r backup.bin
Using clock_gettime for delay loops (clk_id: 1, resolution: 1ns). Found Winbond flash chip “W25Q64.W” (8192 kB, SPI) on ch341a_spi. Reading flash… done (in 1 minute and 4 seconds).
sudo flashrom -p ch341a_spi -E
Using clock_gettime for delay loops (clk_id: 1, resolution: 1ns). Found Winbond flash chip “W25Q64.W” (8192 kB, SPI) on ch341a_spi. Erasing and writing flash chip… Erase/write done (in 3 minutes and 16 seconds).
sudo flashrom -p ch341a_spi -w bios.bin
Using clock_gettime for delay loops (clk_id: 1, resolution: 1ns). Found Winbond flash chip “W25Q64.W” (8192 kB, SPI) on ch341a_spi. Reading old flash chip contents… done (in 1 minute and 5 seconds). Erasing and writing flash chip… Erase/write done (in 58 seconds). Verifying flash… VERIFIED (in 1 minute and 5 seconds). Process completed in 3 minutes and 8 seconds.
I also added a function to disable quick edit on cmd.exe when flashrom is active to prevent accidentally application freeze which might corrupt the flashing process.
On Post #82 it states that the Black CH341A due to improper design supplies the wrong voltage?? Seriously??!! Rather buy a new one than have to solder wires. Guess the green Ch341a is the one to get?
I read however that the jumper placement determines voltage setting such that pin 1-2 is 3.3v and pin 2-3 is 5v?
Been experimenting flashing onto a 25q64bvaig Winbond a P8Z68 Deluxe bios for testing purposes and after the 1.18 ver of software completes its flash the verification fails and says in disagreement…
Yes, it’s important that the supplied voltage on the CH341A (VCC) is identical with the voltage supply on the flash chip.
The green pcb is doing this correctly by suppling 3.3V to all pins:
This jumper only controls the flash chip voltage, the CH341A VCC is always supplied by the USB 5V which outputs it to CS, DO, CLK, DI on the flash chip.
Edit: The jumper controls the CH341A mode (SPI or TTL), only the flash chip VCC is always supplied with 3.3V.
It seems to work correctly with most generic 3.3V DIP8 chips that can be put into the programmer socket. However, it might be dangerous to some flash chips (and more likely to the components around it) if used with a SOIC8 clip adapter.
Edit: Found another possible fix; cut the trace at the pcb back which powers the CH341A VCC with 5V:
Calibrating delay loop… OK Found Winbond flash chip “W25Q16.V” (2048 kB, SPI) on ch341a_spi. Reading flash… done (in 16 sec).
flashrom -E
Calibrating delay loop… OK. Found Winbond flash chip “W25Q16.V” (2048 kB, SPI) on ch341a_spi. Erasing and writing flash chip… Erase/write done (in 33 sec).
flashrom -w random.bin
Calibrating delay loop… OK. Found Winbond flash chip “W25Q16.V” (2048 kB, SPI) on ch341a_spi. Reading old flash chip contents… done (in 16 sec). Erasing and writing flash chip… Erase/write done (in 26 sec). Verifying flash… VERIFIED. (in 17 sec). Process completed in 59 sec.
Did you forget to solder a pin on your wire patch? The guide you linked that detailed the fix had a second solder point that I don’t see in your photo.
That’s for the V3 pin on the CH341A which should be supplied with 3V according to the datasheet, but it seems to work fine without. The green pcb version also doesn’t supply 3V to this pin.
Granted, that’s the whole FOSS community modus operandi, and kind of how the Bus Pirate came to be, but still…I’m seriously appreciative.
