I built all 3 of the floppyemu's today, although the buttons didn't arrive so I was only able to get 2 completely working.
Anyway, here's an FDHD SE with 2 floppyemu's running out of the internal ports and an 800k drive running off the external port. I'm using the latest firmware and booting off a 1.44MB System 6.0.8 install disk image on one of the floppyemus.
Well, dougg3 convinced me I was just being lazy by waiting for the button, and I should just short the pads to test out 3 floppyemus on the SE.
I haven't really investigated much, or tried anything other than the latest firmware, but when there are 2 floppyemus on the internal ports and the Apple 800k drive on the external port, it works fine. But when I replace the Apple 800k drive with a floppyemu, the machine refuses to boot, just repeatedly putting up the X floppy icon.
It happens whether a floppy (800k or 1440k) image is in the external floppyemu or not.
If I disconnect one or both of the internal floppyemus and just have it on the external port, it still gets the X floppy icon.
Since the external floppy port can only do 800k (I think), I'm wondering if it's getting angry at the 1440 support. I'll try to downgrade to pre-1440 support and see if it changes.
That is really funny. Do you have them stuck to the front of the SE with double-stick tape or something like that?
If an image file is read-only on the SD card, it will be mounted as a read-only disk. The next board revision will also support the little lock tab on the SD card, to make everything on the card read-only.
I think there are some bugs in the 1440K disk support. Today while installing System 7.5 from a set of virtual floppies, my Floppy Emu consistently died just after disk 2 was inserted. Sometimes it would lock up, sometimes the screen would go blank, sometimes it would reset back to the version info screen and main menu. I'm not sure how to debug AVR code - I think it requires some extra piece of hardware from Atmel like the Dragon? All the debugging I've done so far has been with blinking an LED or writing to the screen, but that's not going to work if it's crashing in some unexpected place.
Yeah, at the moment they're just stuck on with rolled over electrical tape. I'm trying to brainstorm some ideas for a removable mount from the front of systems like that, probably just some padding on the back of the PCB as a spacer out to the depth of the DB19 connector, then a strip of velcro or something. I can hear the ESD now.
Interesting about the 1440 support. So far, I've mainly been using the 1440 System 6.0.8 floppy images, running TeachText off of them and the like to test keyboard input with the USB adapter on both Plus and SE. So far so good, but that's admittedly a pretty small sample size.
I haven't done any AVR debugging, and I usually do what you're doing, LED and maybe some TTL serial out.
I made a few unrelated (?) bug fixes in the past day or so, and now the crash/reset problem I was seeing with 1440K disks has disappeared. Previously it happened three times in a row while installing 7.5.0 from a set of floppy images, right after I inserted Install Disk 2. But now it works fine! ??? Which is good, because I realized that all the AVR pins are used, so there's physically no place to attach a hardware debugger even if I had one.
I ordered 50 revision 1.1 boards yesterday. Whee! I'll use some to build assembled units, and the others can go to people doing DIY builds, since it's not really economical to order just 1 PCB for yourself.
I took a shot at 1440K write support today. The CPLD is about 100% full, and it took a lot of fiddling to squeeze in the necessary logic for MFM writes (or what I hope is the necessary logic). Unfortunately the first attempt didn't work at all, and the data coming from the Mac doesn't look even vaguely recognizable. I'm going to need to hook up the oscilloscope to see what's going wrong, I think. I'm crossing my fingers it won't be anything too complicated, and won't need any more CPLD logic, as there's virtually no space left. There's some kind of "precompensation" logic that happens for MFM writes, where the bit timings are altered in order to account for anticipated magnetic interplay of two nearby 1 bits on a real floppy disk, so maybe that's causing some problem.
Another mystery is why it doesn't work on Power Macs, when connected to the internal floppy connector. I've tried a Power Mac 7100 and 8500, and it doesn't work on either one. The MOTOR light is constantly dimly illuminated, which is weird. I'm 99% sure the floppy drive in Power Macs is just a regular Superdrive, so I'm not sure why it wouldn't work. Some kind of electrical difference maybe? Or maybe the floppy ROM routines do something unexpected that Floppy Emu isn't designed to handle, though I can't think what that would be.
I think there's 2 different types of Superdrives: a manual inject and a mechanical inject. I'm really not sure if there's any electrical differences between the two here, but just thought I'd mention it. The two different drives have different mounting screw locations, so it's not convenient to use them interchangeably, just from a physical perspective. I don't think I've ever tried swapping just to test the electrical compatibility.
I'll also have to try the floppyemu on the internal port of one of the AV macs or a Performa/Quadra 630, since those are all kind of halfway to PPC.
What's the change for the 1.1 PCB? I was contemplating another PCB order, since that takes the longest to arrive.
Hey bigmessowires, I am interested in buying a PCB or two (for build back-up). I haven't soldered mine yet, so I might as well make one with the latest. Having the backlight LCD and support for SD lock (thinking future school environment use) would be a nice to have. I have most of the parts except for the LCD and floppy connector... I should order those. =)
Sure thing! Can you hang on for another week or two? My hope is to sell a "DIY kit" with the PCB, DB-19 connector, and LCD, since those are the hardest parts to come by and the ones where shipping costs kill you if buying them in individual quantity. I have a big bag of DB-19 connectors now, but still waiting on the other parts.
Bbraun, if you do test out the AV machines, I'd be interested to hear how it goes.
Yeah, I've got 3 working ones, and one almost working one. On the almost working one, one of the pins on the CPLD is shorted to ground, and I can't seem to figure out where/how. Someday I'll get back to it...
I might just wait and buy one of your DIY kits, that sounds pretty handy, and I like the idea of the backlight!
I also didn't populate the ISP header on the ones I've built most recently, now that you have the firmware update feature. I got one of those headers where the pins are larger and tapered so you can just push it into the PCB holes for an easy, good, but temporary connection and did the initial programming that way.
I didn't test on an AV mac yet, but I tested on a Quadra 800, and got some odd behavior.
When trying to boot from the floppy, the ROM doesn't like it and does the X floppy icon. However, the floppyemu doesn't seem to acknowledge the eject. One of the times I attempted this, the floppyemu went to a blank screen and the debug LED was lit.
After booting from a legacy recovery CD, and trying to mount a floppy from the OS, the OS saw the disk was inserted, but thought it was blank and offered to initialize it. I chose to eject it instead, and the floppyemu rebooted.
FWIW, the Quadra 800 should be similar/identical to the Centris & Quadra 610 and 650 in this regard, although I haven't connected it to those to double check.
More bugs! Thanks for testing this. I really need to compile a definitive list of which machines are and aren't compatible. Why do you say the Q800 should be identical to the Centris & Quadra 610 and 650, do they have the same ROM? I've had success with a Classic and a IIsi, and got a success report from someone with a LC475, but no success with a PowerMac 7100. I don't own any machines in the range between those to narrow it down further than that. I should probably get some.