Fantastic work! Recreating old tech shows great dedication. Back in early 2022 I went though the process of reverse-engineering an SX-64 I/O PCB. I was restoring an SX-64 and the I/O board was full of corrosion. Not being able to find a readily available replacement, recreating was the next step. I was lucky that I was able to find a PCB engineer wizard lurking in my amateur radio club :-)
Seems Big Brother Google knows my love for the 264 series (first computers I ever used at age 8), so for once I'm thankful that a new channel was recommended to me, and it was yours with a 1551 video no less. Great work with the drive and the paddle. I have a 1551 in storage that I haven't dared to touch yet, all the knowledge you shared will be a great starting point when I finally dig into it. Thank you!
Device 8 or 9 is selected on the drive (cut a trace between two pads next to the 6525). This is sent via pin 2 of the drive cable. That is used by the PLA to select FEFx or FECx. With no drive connected, that will float, so may explain some of the problems you were seeing on the paddles with no drives connected. If you pull that low, that might make the test program work (or at least be more consistent if the 6523 was faulty).
I reran the test program with the drive cable disconnected and a pull-up on pin 2, and that gave (mostly) correct results (there were occasional 'Bad / unstable' flags, but they weren't consistent between runs). Thanks @DaveCurran
not that i think it matters for the logic test, but the "program from the internet" tests the registers for FECx in every case in the video. i think that program later asks which drive, and then changes the value to FEFx if required. its a bit like that program was written assuming the 1551 drive was going to be an additional drive that was changed to dev 9.
@@fumthings when the drive is powered down, pin 2 will be low, which indicates drive 9, so it is correct to test FECx until the drive is powered on and pin 2 goes to the correct voltage. (my 1551 passes all the tests on that program)
Cart connector continuity was nagging me while watch this until 55:00 Also, check grounding on your soldering station, I have the same model and it came with a loose ground inside. 😬
Sounds like maybe you can modify the paddle to have a IEEE-488 port? That would more or less explain why the cable to the drive is in that staggered "D-sub" shape.
You probably could, but the protocol is not IEEE-488, so you'd also need to rewrite the DOS. At which point it might be simpler to convert a 1541 (there are conversion PCBs & ROMs available for that)
Do you have a usb 3.5 inch floppy drive? The Pi1541 is supposed to have experimental usb support for thumb drives, but I don't think anyone has experimented with a usb floppy drive
@@TimsRetroCorner unlike 5.25 disks, 3.5 inch disks seem to be surviving. Might be an affordable method for those of us who want to use real floppies. Any ideas?
Great job! When you designed a new board, why didn't you design it around the 27512 EPROM? That is a lot easier to find and make than getting a PLA chip.
@@TimsRetroCorner That PLA can't be that easy to get either or is in short supply. For the C64, two GAL chips are used to replace the PLA (I guess more or less the same as an EPROM, just a cross or table. That might be something to try to breadboard for all 10 of these drives that remain out in the wild - lol.
@@Daveyk021 Wasn't that hard to find (as a programmed up PLS100), just a bit pricey is all. Once I've checked the EPROM I made works (shouldn't have been so quick to put back the cartridge shield, lol), when I lay out the Mk2 board, I can add an eprom footprint and give people the option of either.
One of the first products I worked on used that PLS100 chip back when it was still a Signetics product - admittedly, we were using it as a state machine and running in on a fast clock, but it quickly got a reputation from the service people as a "replace on sight" part. I've been sort of biased against them ever since.
Yeah, MOS had a lot of problems with passivation, especially early on. For example, the MOS 6550 1024x4 SRAM that was used in the original PET was especially bad.@@TimsRetroCorner
More likely greedy sellers on fleabay gouging the prices of original chips. People that do actual work to produce things are worth paying for their time and effort and skill. And that includes reverse engineering stuff.
@@TimsRetroCorner It is a pitty I don't have 1551. I can do the reverse engineering and do it opensource. :) And I just looked in bay... the only available is 300+ wow!
I have one (you can see it on my bench at various points in the video), but I rarely use it because it's always clogging up and a right PITA to clean. Maybe a genuine Hakko would work better, but those are expensive!
@@TimsRetroCorner Interesting I didn't have that issue with mine. And I just did a capjob of a Mega Drive one with it... However, I added a bend, holey Metal Plate from my old Xytronics LF8800 to the ZD915. (for the price, its pretty bad due to the weak pump) Maybet hat's why it didn't clog? At least not nearly as much as the Xytronics...
Fantastic work! Recreating old tech shows great dedication. Back in early 2022 I went though the process of reverse-engineering an SX-64 I/O PCB. I was restoring an SX-64 and the I/O board was full of corrosion. Not being able to find a readily available replacement, recreating was the next step. I was lucky that I was able to find a PCB engineer wizard lurking in my amateur radio club :-)
Modern PCB fabs and tools are so cheap and accessible, I'd never have been able to do that a few years ago!
Seems Big Brother Google knows my love for the 264 series (first computers I ever used at age 8), so for once I'm thankful that a new channel was recommended to me, and it was yours with a 1551 video no less. Great work with the drive and the paddle. I have a 1551 in storage that I haven't dared to touch yet, all the knowledge you shared will be a great starting point when I finally dig into it. Thank you!
Great work, buddy! I always wanted to see the workings of a 1551 drive, thank you very much!
What a fantastic job you just did!
ABS can be solvent welded back together it can also help messed up threads by mushing it all around. 38:00
Wow, this is tremendeus amount of work on this interface, nice job :) Cheers!
Thanks. I still have to build a case for the new interface (as well as make some tweaks to the pcb)....
Great job on this tim glad it worked out in the end
Wow that was really impressive work Tim. I have a couple of Plus4 and C16's, but I never hade a 1551. Interesting video.
This is my first 1551, so of course I was determined to get it working!
How cool. Great Job on the pcb. I've always wanted to find a 1551 for my setup
Now that's really going the extra mile! Great job!
Great job!!! What a mare that was to get working lol
Cool Project and great work! 👍
Device 8 or 9 is selected on the drive (cut a trace between two pads next to the 6525). This is sent via pin 2 of the drive cable. That is used by the PLA to select FEFx or FECx. With no drive connected, that will float, so may explain some of the problems you were seeing on the paddles with no drives connected. If you pull that low, that might make the test program work (or at least be more consistent if the 6523 was faulty).
Because it acts as an address input on the PLA, and while it is pulled up, it's pulled up inside the drive. Yes that makes sense.
I reran the test program with the drive cable disconnected and a pull-up on pin 2, and that gave (mostly) correct results (there were occasional 'Bad / unstable' flags, but they weren't consistent between runs). Thanks @DaveCurran
not that i think it matters for the logic test, but the "program from the internet" tests the registers for FECx in every case in the video. i think that program later asks which drive, and then changes the value to FEFx if required. its a bit like that program was written assuming the 1551 drive was going to be an additional drive that was changed to dev 9.
@@fumthings when the drive is powered down, pin 2 will be low, which indicates drive 9, so it is correct to test FECx until the drive is powered on and pin 2 goes to the correct voltage. (my 1551 passes all the tests on that program)
@@DaveCurran ah, thanks, that all makes sense now.
Great video Tim! I have a faulty 1551 too. Maybe i can pin the bug down now with what you showed here..
Cart connector continuity was nagging me while watch this until 55:00
Also, check grounding on your soldering station, I have the same model and it came with a loose ground inside. 😬
Will there a FPGA version of the chip that went bad.
I'm not aware of one being developed.
I see I'm not the only one with a 1551 in bits on my desk this week. Interesting.
Sounds like maybe you can modify the paddle to have a IEEE-488 port? That would more or less explain why the cable to the drive is in that staggered "D-sub" shape.
You probably could, but the protocol is not IEEE-488, so you'd also need to rewrite the DOS. At which point it might be simpler to convert a 1541 (there are conversion PCBs & ROMs available for that)
@@TimsRetroCornerSo in some weird way they made up *another* protocol besides 488 and their own serialized version of 488?
How much faster is the 1551 compared to the 1541 without fast load?
Haha, you already answered that in the end of the video! I asked too quickly :)
lol
@@TimsRetroCornerHow does it compare to the C128/1571 burst mode transfer?
Do you have a usb 3.5 inch floppy drive? The Pi1541 is supposed to have experimental usb support for thumb drives, but I don't think anyone has experimented with a usb floppy drive
Funnily, I was just thinking about that in another context. It'd be fun to try...
@@TimsRetroCorner unlike 5.25 disks, 3.5 inch disks seem to be surviving. Might be an affordable method for those of us who want to use real floppies. Any ideas?
Great job! When you designed a new board, why didn't you design it around the 27512 EPROM? That is a lot easier to find and make than getting a PLA chip.
Because I had the PLA, and I had no proof at the time that the EPROM solution worked.
@@TimsRetroCorner That PLA can't be that easy to get either or is in short supply. For the C64, two GAL chips are used to replace the PLA (I guess more or less the same as an EPROM, just a cross or table. That might be something to try to breadboard for all 10 of these drives that remain out in the wild - lol.
@@Daveyk021 Wasn't that hard to find (as a programmed up PLS100), just a bit pricey is all. Once I've checked the EPROM I made works (shouldn't have been so quick to put back the cartridge shield, lol), when I lay out the Mk2 board, I can add an eprom footprint and give people the option of either.
@@TimsRetroCorner "just a bit pricey"... show that equations! Make it opensource and everybody can program any programmable thing to emulate PLA.
One of the first products I worked on used that PLS100 chip back when it was still a Signetics product - admittedly, we were using it as a state machine and running in on a fast clock, but it quickly got a reputation from the service people as a "replace on sight" part. I've been sort of biased against them ever since.
same thing now with MOS branded logic.
Yeah, MOS had a lot of problems with passivation, especially early on. For example, the MOS 6550 1024x4 SRAM that was used in the original PET was especially bad.@@TimsRetroCorner
I think, that PLA must fit to a bit more "modern" chip without problem. The only problem, that greedy people hide information :)
More likely greedy sellers on fleabay gouging the prices of original chips. People that do actual work to produce things are worth paying for their time and effort and skill. And that includes reverse engineering stuff.
@@TimsRetroCorner It is a pitty I don't have 1551. I can do the reverse engineering and do it opensource. :) And I just looked in bay... the only available is 300+ wow!
You really should look for a ZD-915 and invest in it.
It would make your life much easier and its not very expensive, over here it was ~100€ shipped.
I have one (you can see it on my bench at various points in the video), but I rarely use it because it's always clogging up and a right PITA to clean. Maybe a genuine Hakko would work better, but those are expensive!
@@TimsRetroCorner Interesting
I didn't have that issue with mine.
And I just did a capjob of a Mega Drive one with it...
However, I added a bend, holey Metal Plate from my old Xytronics LF8800 to the ZD915.
(for the price, its pretty bad due to the weak pump)
Maybet hat's why it didn't clog? At least not nearly as much as the Xytronics...