ijor

Btw, the lack of SIO audio In has an interesting side effect (besides actually loosing the tape audio channel). You don't hear the disk loading audible crosstalk when Pokey volume is off. You could simulate the sound, I guess. Not saying you should, just thinking in loud voice

Interesting. Did you research why it didn't work? May be something to do with the eeprom writer?

Useful info, thanks. But my problem was that I did it on Windows 7 that doesn't have the USB driver. Otherwise it wouldn't have been much of a problem. Actually Xloader from the link I posted above it does work quite well and makes the whole process very simple and quick. That is assuming you do have the USB driver and you don't want to build the code yourself, just flash. Btw, I forgot that I have a clone Atmel AVR programmer that I bought a few years ago but I never used it. Guess I could use it to flash the AtMega directly bypassing the USB altogether. Not sure it is really worth except to test the programmer

Well, sort of. There are some issues and probably it won't work with every single title. The main problem is that most titles are not aligned to the index hole. This requires that the software would perform a high level analysis and would find a proper write splice point for each track. The SCP is better for writing purpose precisely because we already have software (Phaeron's a8rawconv) than can perform this task. As commented above, there is also an issue when writing a disk with an 80 tracks drive and reading it with a 40 tracks one as a 1050 (or a 810).

You don't connect it to the XF551 itself. You connect it to the drive that is inside the XF551, which is a standard PC drive mechanism. Depending on the cables all you might not need to disassemble the XF551, just remove the top cover and disconnect the the mechanism to the internal XF551 board.

The Blue Pill is not buffered, that's what you get for something that costs ~U$1.50. But it should work with 5.25 HD drives, and probably even with not so old 5.25 LD drives as well. You can still add buffers if you want yourself for very little cost. I would prefer something safer and more robust. But the whole point of this is a very cheap DIY alternative for those that can't afford the commercial boards.

Ok, if you insist First problem was to install the arduino USB driver that is required to flash the device over USB. Note that I never do that kind of stuff in my main machine. I used a system running Windows 7. I downloaded Xloader from this page: https://atari8bit.net/de-re-sdrive-max because it was supposed to automate the whole process installing the drivers among other things. Well, it didn't install any drivers. So as recommended on other pages I downloaded the whole Arduino IDE which is supposed to have drivers for every type of Arduino, or so I thought. I didn't install the IDE, I just unzipped it and pointed Windows to search for a driver in the Arduino folder, but it couldn't find anyone matching the Sdrive-Max. So I decided to went along and actually install the Arduino IDE and let it install every driver it asked me during the process. I would never do something like that in my main machine, of course. Even then, after full install and reboot, the Sdrive-Max was still unrecognized by Windows. It was at this point that I checked the properties of the USB device and looked up the USB VID/PID (the USB device ID) and performed some research. It turns out that the USB interface in my Arduino is made by some obscure Chinese manufacturer. Windows 10 can automatically detect the device and install the corresponding driver, but for some reason Windows 7 doesn’t and can’t even find any driver online. Seems the only way is to download the driver manually from some Chinese web page. And of course that it took some time to find out all of this. So I had to browse and navigate a web page in Chinese only, guess what file I need to download, and more important, guess which one is the right button to initiate the download. I hoped that the Chinese text I clicked meant “download” and not something like "Yes, I agree to participate in some kind of coronavirus trial" Finally I had the right driver installed and Windows could recognize the Max and I could use Xloader to flash the new firmware. But as I said before, I didn’t know which LCD controller I have and then didn't know which version of the firmware to install. As hinted by The Doctor, indeed the Max flashes the current LCD controller for a fraction of a second at boot time. I seemed to read ILI9340 (guess I’m getting older ...) and tried to flash that version. The firmware upgrade worked, at least apparently as I could calibrate the touch and the Max booted the new firmware ok. So I thought that I flashed the correct version. But then when inserting the card and trying to actually use it, it didn’t work well. I thought the problem might be something particular to the latest version not being fully stable. So I downgraded to an older version, but now even the touch calibration wouldn’t even work. It was only at this point that I realized I might have flashed the version with the wrong controller. It didn't occur to me before, because I thought that in that case it wouldn't work at all and it would show nothing on the LCD. I guess that the reason that the latest version worked partially, even when using the wrong controller, it's because it has some kind of auto detection? Lastly I flashed the last version with the ILI9341, as it seems that is the most common LCD controller, and it also turned out that this was the one I actually have. And then, a couple of hours later and a handful of flash erase cycles wasted, my Max is finally happy

Useful post. Thanks. The "F7 v2" might be the cheapest version that is sold fully assembled, at least by Keir. But note that you can build yourself a bare version based on the Blue Pill for close to nothing. The Lightning version, at £40, doesn't seem to me so cheap anymore. Not much cheaper than the original SCP, and if you live at the US and you have to add international shipping, the Lightning might cost about the same or even more.

Re: Firmware upgrade Thanks everybody for the comments. I finally could upgrade the firmware to the latest. I won't bore you with the details, but it took some time and it wasn't easy.

Firmware upgrade: I am trying to upgrade the firmware but I don't know which LCD controller I have (I didn't build it). Is there anyway to know which firmware variant was flashed before? Or should I just try to reflash the firmware for every controller one by one, until one that works? Probably I am missing something here: Last version is supposed to have autodetection? Then why there are still multiple versions for each LCD controller? Thanks,

Hi ggn, how are you doing? Note that if you want, the mechanism inside the XF551 is probably good enough. They are not among the best 5.25 drives, but should work. Please note everybody that both 40 and 80 tracks work as well for reading purposes. As always, and not specific to these devices, there is an issue when writing a disk on 80 tracks drive and then reading the disk on a 40 tracks drive like a 1050.

Make sure it's the same, or equivalent, to the "plus" version I linked above. The standard version is made using the Blue Pill board and it is extremely cheap. The Greaseweazle wiki mentions a version made with a custom board that uses a more powerful version of the MCU (F7 instead of F1), but otherwise it seems to be the same. I don't know if they call that version "plus" and if that's what they are selling on facebook or not. Then there is the "plus" version I mentioned that besides the upgraded MCU, it also adds buffers and a few more thing. I don't know if it is worth just the upgraded MCU. It is IMHO overkill for this task. Of course, if it costs about the same then it doesn't harm. But I think the safer interface with buffers and ESD protection is much more important.

I made a typo when creating the thread. It's SCP, not CSP. Can any moderator change the title of the topic, please (and delete this post if you want)?

I thought we were already hijacking this thread too much and I opened a new thread on the main subforum. Yeah, it is still cheaper than the Kryoflux or the SCP. Part of the extra cost is probably that uses a different MCU that is much more powerful. It uses the STM32 F7 variant that, IMHO, is completely overkill (~500MHZ DMIPS plus FPU).

We talked about this on one of the subforums, but it was a bit hijacking another thread. And I thought this deserves his own thread here. There are (were) currently two main devices to perform preservation/flux level dumps of floppy disks, the Kryoflux and the SCP. They are both good and recommended devices, but many people find them too expensive. Since about some time there is a much affordable alternative that is DIY and fully open source: https://github.com/keirf/Greaseweazle/wiki It is based on the extremely cheap "Blue Pill" STM32 development board: https://stm32-base.org/boards/STM32F103C8T6-Blue-Pill.html Here are a few comments, but note that I don't have personal experience with this board: - I don't know much was tested with Atari 8-bit disks if at all - When dumping disks it produces SCP files as the original SCP hardware does - I don't know if it supports the mod to be able to read the flippy side as the Kryoflux does - Writing should work about the same as using a Kryoflux. The SCP is currently better for writing because it is supported by Paheron's a8rawconv that can setup a suitable write splice point. - The floppy interface is completely unbuffered that might not be suitable for very old 5.25 drives Somebody sells a ready made "plus" version that among other things it adds buffers, ESD protection and also uses a much more powerful MCU. But then the cost might be no so cheap anymore: https://www.sellmyretro.com/offer/details/greaseweazle-f7-plus-usb-floppy-interface-flux-tool-usb-_b-43881 Would be interesting to hear if somebody tried this with Atari 8-bit disks.

Try removing the disk image, leaving the Gotek drive "empty", before inserting the other one. Or temporarily swap to a disk image file that has the opposite read only status. The ST has a particular disk change detection based on the disk write protection. The exact implementation depends on the specific TOS version and it doesn't work with most PC drives. It is possible that the Flash Floppy doesn't exactly emulate an original ST drive. But even then, here it is worse because you change disk instantaneously, something that you can't do with real disks. There should be here in the forum several threads describing the issue in detail.

Just a bit of a warning about the Greaseweazle. One thing I don't like too much is that the floppy interface, at least in the original version, is completely unbuffered which is not very robust. With very old drives it might not work and it might even damage the MCU (the main processor chip). That won't be the end of the world anyway because the Blue Pill that comes with the MCU is so cheap. It shouldn't be much concern as long as you use a not so old, 5.25 " HD 80-tracks drive. Just double checked in case this was fixed ... There is somebody that sells a "plus" version that adds buffers and ESD protection. But then the cost is no so cheap anymore: https://www.sellmyretro.com/offer/details/greaseweazle-f7-plus-usb-floppy-interface-flux-tool-usb-_b-43881

Yeah Don't know if it's a coincidence or not. What I can tell you is that they are certainly developed by different people.

But seriously, and once again, both you and Mark did a wonderful job here!

The SuperArchiver (standalone) and the BitWriter work at completely different levels. I don't have much experience with the BitWriter. Yes, it works at a lower level than enhancements like the SA or the Happy that sill depend on the FDC capabilities and limitations. The BitWriter is not a flux transition level device but it bypasses the FDC. I'm not sure how reliable it is, though. And because it uses a standard 1050 mechanism it still lacks the accurate index position. Note that there are (or were) a few other flux level devices, like the Happy Discovery Cartridge for the ST. And btw, there is a new device that is DIY and completely open source: https://github.com/keirf/Greaseweazle/wiki I'm not sure if somebody tried it with 8-bit disks, and I don't know if it supports the modded drives to easily read the flippy side. Writing might be problematic, but besides the flippy issue it should work for reading, at least with most drives. And if you build it yourself it is probably very cheap so not much harm is done trying. One more thing about imaging original disks with 1050 enhancements. As noted in may other threads, please be careful as some originals are extremely fragile and they might even tear off apart when attempting to read them. This could happen also when using a SCP or a Kryoflux. But it is worse, in this sense, with 1050 enhancements because it needs more revolutions per track, which means the fragile surface is "suffering" the friction even longer.

Very useful information, thanks Jeff. I have a 1200XL that is brand new, or at least in brand new condition. The keyboards was completely dead. No key worked at all, not even the Reset key. I followed Jeff instructions, disassembled the keyboard, found the mylar looking pristine and connected the naked pcb. Indeed it works fine without the actual keys. I would prefer to avoid messing with the springs. So I reconnected the whole keyboard, tried the "tjlazer" method and started working out the keys. After many hits, a few keys started to come alive! Still a long way to go, it feels like this keyboard will need something like 500 hits per key, not just 50. But, oh well, things we do for Atari!

Still down, too bad ... Anybody heard anything from Dal? Any news at the Facebook group? Thanks,

Hi Mark, Forget about Pokey, forget about the core. I am talking about the pure Eclaire board hardware. I can see that motor control is buffered, so there is no way to read its state (it's a pity). But the rest of the signals don't seem to be buffered and are fully bidirectional and I could reverse their direction if I wanted it, right? Also, again, did you test connecting a tape recorder like a 1010? Or if I try I would be the first with the inherent risk of blowing the whole thing?

A couple of questions ... Somebody tested the SIO external interface with a 1010 or other Atari tape recorder? I understand that the whole SIO interface, except motor control, is bidirectional, from the hardware point of view, correct? In theory I could reverse the direction of any SIO signal (again, except motor control), if I wanted, right?

A SuperArchiver, or Happy or similar enhancements, are not really suitable for preservation. It won't work with all titles, and for those it will, it won't be very precise. It is ok if for some reason you want to make your private copies of your own disks. It is also acceptable, even for preservation, if you have some really ultra rare stuff (because any kind of dump is better than nothing) and you live on the moon where you can't contact somebody nearby to get those disks preserved properly with a SCP, Kryoflux, or other low level solution. I don't want to discourage you to make digital copies of your disks. Certainly not, after all I also started with a Happy many years ago. It might be even fun. Just be aware that this is not the best preservation method to say the least.