Hey, thanks for featuring my board on your video. Nice build. To speak to your case wobbling, in my experience, this happens if the 3D printer that was used to print the case isn't perfectly square. The resulting print is then slightly out of square, and when assembled, it twists ever so slightly and causes it to rock. IF that is the case with your print, then you might want to mention to your friend to check their printer for square. On the topic of enhancements, I really like your suggestions, and in all honesty, some of those had been a passing thought subsequent to my last revision. I think I'll take those suggestions to heart and make a version 3. I do really like your thoughts on the double FD pin header to allow for either type of drive to be plugged in. I think removing the resistors will leave plenty of room for the circuit and IC. And lastly, on the thoughts on an FD-2000/4000 board... that's in the works!
This is fantastic! I'm looking forward to your revisions and the FD-series project! The case was a rough print, but it turned out much better than I expected so I used it as-is. The wobble artifact may very will be a side effect of the setup as you suggested, and I'll let my friend know.
Love how the Commodore community is "somewhat" discovering the DIY aspect of the hobby. Considerably later because they used to stick with "that can't be done" for way too long, but hey maybe the future might yield something making me return at some point, even tho I consider that very unlikely.
500 USD nowadays for a working 1581? OMG. I used to have one in 1994, bought it with a defective disk drive, paid like 15 Deutsche Mark, and put an old Amiga Chinon drive in that I had laying around and boom it worked. Sold it in 2005 and by then the price already was like 50 Euros (100 Deutsche Mark, 55 USD). But 500? I should have kept it :( and several other vintage items I sold back then, too.
"...And the through-hole component is what makes that possible." Not necessarily just possible, but reasonably easy in comparison. Surface-mount components _can_ be hand-soldered but they'd probably be a huge PitA to most people who would even try this!
Just give SMD a chance and you'll very likely never look back - it's at least what happened to me and a few friends. Through hole is just annoying with the constant board flipping, alignment and snipping off excess component leads. It's incredible how most retro communities are cool with SMD, except for the 8bit Commodore one which has always been extremely resilient.
You might be able to fix the case "creaking", depending on what the problem is. The two longest screws that are used to hold the halves together go through a long plastic "tube". If you shorten the length of the tubes (filing) so the tubes don't physically touch the bottom (I'm assuming they're long enough to touch) you might be able to put a little extra force on the screws to force the case halves together. If there's too much force needed to squeeze it together, use small o-rings sandwiched between the plastic tubes and the bottom. The o-rings will squash like rubber and provide a flexible compliance between the attached halves. You might have trim the length of the plastic tubes as needed for a snug fit.
Very nice video, I have built many 1581 for friends and good known. The main problem are the floppies themselves you can only take DD 720KB floppies, but if you put a 1.44 MB HD floppy disk the drive does not work, even a masking of the HD hole is not a good idea for data retention. DD floppy disks are becoming less and less common but HD floppy disks are still enough, you would have to modify the drive that it can also safely write to HD floppy disks - test it.
That is froogin' awesome dood. There is a replacement switching mini power supply out there for $18-$40. On Corei64's design you can replace the 7805 with a TSR-1 2450 and not have to buy the heat sink. Also takes up a lot less space.
Very interesting and thought provoking presentation, Steve. I'm currently collecting components for the DIY Chris board, but your mentioning the other PCBs has given me pause for thought. I particularly like the idea of a dual unit (although I wonder if there's a way to configure them as (say) Unit 8 drives 0 and 1 -- like the old PET drives. I'm sure I read that the code for it was still in the 1571 roms at least, so maybe it's in the 1581 too?)
Thank you for the feedback, Tim! I wasn't aware of the device 0/1 code still lingering around in the ROMs -- that would be a nice addition to the dual 1581 drive. Perhaps Thomas from the 8-bit Resurgence channel has more info on the subject.
I got the board with the regular on it. It looks like to me if you run a jumper wire from the regular to the the 12v line and you can power this with the DIN with just 12v. I agree with removing the resistors and use that room to have 2 headers.
Yup, I think that will work. The 2pin header that serves as input to the voltage regulator is a good place to tap the 12V from the DIN port. Also, keep in mind the jumper that determines where the board's 5V rail is sourced (DIN or regulator).
@@SteveGuidi I am running it off 12v it works fine. but the 7805 gets HOT! so I ended up ordering up some 5v buck convertors that are 7805 compatible. they where cheep on Ebay $13 for 10 The lights where way to bright for me. so I ended up putting a 3K in to R13 and a 1K in to R12.
I have a couple ideas to power the drive from a USB-C power supply with Power Delivery. In practice, we only need 5V at about 2-3A, but using 12V and a small buck regulator will make a tidy and compact versatile supply.
Are you referring to the rocking/creaking noise problem with my case? In this situation, it is the top of the case that is misaligned with the bottom part. One of the posts may have a printing imperfection, or maybe I'm using screws that are too large.
Just wondering: could you solder a NEC UPD 765 with supporting logic on a board that sticks right into the expansion board. That way you can hook up 4 3,5 inch shugart drives from a single 34p flat cable, instead of building a complete 6502 computer for each drive.
hey, i know which chips to use but what pieces need to go in there, like diy chris has a list to buy them but they are expensive like what resistors, what crystal oscilator, and that hole other stuff?
So are you saying that you might be saving a whole whopping $7 or so by when you build your power supply that you were going to build anyway, just including the cheaper switch in that? Why go out of your way to do all that workaround stuff just to save around $7 instead of just giving in and having a proper period-fit switch right in the external drive?
"Dual 1581..." * 1582, probably. Oh, wait, that would probably be if you're using drives 0 and 1. But it sounds like you're talking about a dual- *device* 1582 or something like that. But Commodore's standard would've had them as drives 0 and 1 in the _same device_ (8, 9, 10, or 11), used in this format: ' "0:"..." ' or ' "1:"... '. With the way most devs programmed for disk drives, either setup would normally be pretty useless, unfortunately. That sort of expansion was VERY underused for the 20 and forward, and that's too bad. It would be interesting, though!
I assembled this whole thing, but all I'm getting in the red light blinking three times. I'm not sure what is going on here. I've tried everything I can think of. Different SRAM chips, different CPU, different EPROM's.
I am in a similar situation; the serial bus hangs when I try to read it and the computer needs a reset to recover. I'm not sure what's wrong, but I'm waiting for a friend to finish his builds and produce a working device. Once that's done, we can compare part and swap components, etc...
@@SteveGuidi Same thing is happening to me. Serial bus hangs as well and the computer is frozen. Can't read any meaningful error messages. Turn the drive off, the computer unfreezes.
@@aa7aeThank you for the suggestion! I never considered that the SRAM might be a problem as it was purchased new from DigiKey. I'll have to test that out!
@@SteveGuidi I found out my SRAM is all good after all. My Retro Chip Test Pro was not working right because I had soldered in the wrong value resistor for the MOSFET. The self test kept failing which lead me to troubleshooting. Once that was corrected, now all my SRAM tests good. I thought I had a smoking gun, but now it would seem I'm back to square one again. I'm going to test all the logic IC's but I doubt any of them are bad. Whatever is causing this built not to work is quite a mystery at this point.
I got a message back from the board supplier in which he shared experiences where using modern equivalents of the logic IC, 7406 and 7407 may not work in some cases (U8, U11, and U12). E.g. Texas Instruments version commonly available on Digikey and Mouser. Problems seem to go away when using older NOS Fairchild variants. I'm am sourcing some of those on EBAY from U.S. Seller and will see if that makes a difference. The 1581 service manual describes the parts of 7406 and not 74LS06N. I would think they should be equivalent, but maybe the older IC's behave slightly different? We'll try and see if this solves the problem.
The 1581 is quite a powerful totem artifact in the 8 bit C= community. Good thing I have two of them. I'm not here to gloat, but I do have two 1581s. :) Nice project BTW.
This is an expensive project to be sure, and I really have to wonder about how things have changed in the year since you made this video both in terms of making this project perhaps way more expensive and in terms of making an analogue potentially a lot cheaper. What I mean is … 3.5" floppy mechanisms are still common enough, but I wonder if some of the chips you need for this kit are just no longer out there in quantity enough to make this viable. Something similar happened with the BYO VIC-20 kit. Made with all new parts … except the MOS 6560 of course. The VIC chip. At the time they were cheap and common enough on eBay for the intended use-replacement parts to bring a dead VIC-20 back to life. Not enough for a few hundred hobbyists to buy them all up hoping to build kits! If you were designing a kit like that with the scarcity of those chips now (they sell for > $100!) maybe you'd start by assuming those parts are unobtanium and try to get one of the more powerful microcontrollers to emulate it. The RA4M1 (Arduino Uno R4) can handle 5v logic, but it's SMT soldering so maybe you want to just have pins for the Arduino to plug in. You're gonna get people calling it "cheating", but it's literally emulating one chip, so they can stick it. Or if you're willing to do a little more to interface with the 6502 bus, the RP2040 or STM32 are even cheaper options. These are all SMT chips unless you get them with their common implementation boards, so that violates one of your rules of sticking to through-hole unless you make other more significant changes to the design. That said, if you're replacing a chip with a controller that can do more complex logic … can it also do some trivial glue functions without feeling like you're "cheating" too much? And is it worth "cheating" to bring the complexity and more importantly the cost down quite a bit without breaking backwards compatibility? It doesn't look like anyone's done this kind of thing yet, but … if prices for things continue to climb, I imagine it's only a matter of time. Maybe that will truly make it a 1581 for all? Maybe.
Fortunately, the most expensive IC (MOS8520) in the trio of no-longer-manufactured parts is still around $20 on eBay. I bought a WD1770 a few weeks ago and it was a bit more expensive than when I first bought it, but nothing drastic (about an extra $2-$3). So it looks like prices are somewhat stable for this project. That said, the Foenix FNX1591 is a new 1581 clone drive that uses an FPGA at its core along with many SMD components. I believe the only "old" chip needed is the WD1770, but there are plenty of those in supply. Not a kit and the price is close to what you would pay for a 1581 kit. c256foenix.com/product/fnx1591/
Hey, thanks for featuring my board on your video. Nice build.
To speak to your case wobbling, in my experience, this happens if the 3D printer that was used to print the case isn't perfectly square. The resulting print is then slightly out of square, and when assembled, it twists ever so slightly and causes it to rock. IF that is the case with your print, then you might want to mention to your friend to check their printer for square.
On the topic of enhancements, I really like your suggestions, and in all honesty, some of those had been a passing thought subsequent to my last revision. I think I'll take those suggestions to heart and make a version 3. I do really like your thoughts on the double FD pin header to allow for either type of drive to be plugged in. I think removing the resistors will leave plenty of room for the circuit and IC.
And lastly, on the thoughts on an FD-2000/4000 board... that's in the works!
This is fantastic! I'm looking forward to your revisions and the FD-series project!
The case was a rough print, but it turned out much better than I expected so I used it as-is. The wobble artifact may very will be a side effect of the setup as you suggested, and I'll let my friend know.
Love how the Commodore community is "somewhat" discovering the DIY aspect of the hobby. Considerably later because they used to stick with "that can't be done" for way too long, but hey maybe the future might yield something making me return at some point, even tho I consider that very unlikely.
500 USD nowadays for a working 1581? OMG. I used to have one in 1994, bought it with a defective disk drive, paid like 15 Deutsche Mark, and put an old Amiga Chinon drive in that I had laying around and boom it worked. Sold it in 2005 and by then the price already was like 50 Euros (100 Deutsche Mark, 55 USD). But 500? I should have kept it :( and several other vintage items I sold back then, too.
"...And the through-hole component is what makes that possible."
Not necessarily just possible, but reasonably easy in comparison. Surface-mount components _can_ be hand-soldered but they'd probably be a huge PitA to most people who would even try this!
Just give SMD a chance and you'll very likely never look back - it's at least what happened to me and a few friends. Through hole is just annoying with the constant board flipping, alignment and snipping off excess component leads. It's incredible how most retro communities are cool with SMD, except for the 8bit Commodore one which has always been extremely resilient.
@@1337Shockwav3: Ha, I already did. It's too small for me to handle right. Surface-mount is for CNCs/robots and ovens; not me.
Correction: These devices were NOT popular in the 90s. They were popular in the late 80s. The world have moved on from 8-bit C64s by the 90s.
When I discovered they existed, it was already the early 90s, but after all I was still under 10 years old :P
You might be able to fix the case "creaking", depending on what the problem is. The two longest screws that are used to hold the halves together go through a long plastic "tube". If you shorten the length of the tubes (filing) so the tubes don't physically touch the bottom (I'm assuming they're long enough to touch) you might be able to put a little extra force on the screws to force the case halves together. If there's too much force needed to squeeze it together, use small o-rings sandwiched between the plastic tubes and the bottom. The o-rings will squash like rubber and provide a flexible compliance between the attached halves. You might have trim the length of the plastic tubes as needed for a snug fit.
Changing the ttl glue logic to surface mounts would save a lot of space.
Cool video! Happy Christmas (2023)!
A 1581 enhanced board that fits in a 1541 case would be nice too.
Very nice video, I have built many 1581 for friends and good known. The main problem are the floppies themselves you can only take DD 720KB floppies, but if you put a 1.44 MB HD floppy disk the drive does not work, even a masking of the HD hole is not a good idea for data retention. DD floppy disks are becoming less and less common but HD floppy disks are still enough, you would have to modify the drive that it can also safely write to HD floppy disks - test it.
That is froogin' awesome dood. There is a replacement switching mini power supply out there for $18-$40. On Corei64's design you can replace the 7805 with a TSR-1 2450 and not have to buy the heat sink. Also takes up a lot less space.
Based on a 1kg roll of PLA costing $20, the case would have cost approximately $7 in materials.
Damn, an amiga fpga board on it... what a form factor!
Very interesting and thought provoking presentation, Steve.
I'm currently collecting components for the DIY Chris board, but your mentioning the other PCBs has given me pause for thought. I particularly like the idea of a dual unit (although I wonder if there's a way to configure them as (say) Unit 8 drives 0 and 1 -- like the old PET drives. I'm sure I read that the code for it was still in the 1571 roms at least, so maybe it's in the 1581 too?)
Thank you for the feedback, Tim! I wasn't aware of the device 0/1 code still lingering around in the ROMs -- that would be a nice addition to the dual 1581 drive. Perhaps Thomas from the 8-bit Resurgence channel has more info on the subject.
I got the board with the regular on it.
It looks like to me if you run a jumper wire from the regular to the the 12v line and you can power this with the DIN with just 12v.
I agree with removing the resistors and use that room to have 2 headers.
Yup, I think that will work. The 2pin header that serves as input to the voltage regulator is a good place to tap the 12V from the DIN port. Also, keep in mind the jumper that determines where the board's 5V rail is sourced (DIN or regulator).
@@SteveGuidi I am running it off 12v it works fine. but the 7805 gets HOT! so I ended up ordering up some 5v buck convertors that are 7805 compatible. they where cheep on Ebay $13 for 10
The lights where way to bright for me. so I ended up putting a 3K in to R13 and a 1K in to R12.
Your PSU is so small, it could even fit within the case, wouldn't that be beautiful?
I have a couple ideas to power the drive from a USB-C power supply with Power Delivery. In practice, we only need 5V at about 2-3A, but using 12V and a small buck regulator will make a tidy and compact versatile supply.
Perhaps just add some rubber-feet where you can file away some of one to align it better and make it more flat in its contact surface?
Are you referring to the rocking/creaking noise problem with my case? In this situation, it is the top of the case that is misaligned with the bottom part. One of the posts may have a printing imperfection, or maybe I'm using screws that are too large.
Just wondering: could you solder a NEC UPD 765 with supporting logic on a board that sticks right into the expansion board. That way you can hook up 4 3,5 inch shugart drives from a single 34p flat cable, instead of building a complete 6502 computer for each drive.
hey, i know which chips to use but what pieces need to go in there, like diy chris has a list to buy them but they are expensive like what resistors, what crystal oscilator, and that hole other stuff?
The bill of materials and costs are linked in the video description.
Thanks
pi1541 baby!
If there is the STL that includes the top vent slits, then why did you not use that one? Also: Why black instead of beige filament?
"When the case was first printed out..."
How many times did you supposedly print it later?
Your friend doesn't know how much filament that print used and how much the filament cost per length unit?
"Other additional..." or just "additional"?
"Predrill or drill a pilot hole"?
What's the supposed "difference," according to you?
So are you saying that you might be saving a whole whopping $7 or so by when you build your power supply that you were going to build anyway, just including the cheaper switch in that? Why go out of your way to do all that workaround stuff just to save around $7 instead of just giving in and having a proper period-fit switch right in the external drive?
"Dual 1581..." * 1582, probably. Oh, wait, that would probably be if you're using drives 0 and 1. But it sounds like you're talking about a dual- *device* 1582 or something like that. But Commodore's standard would've had them as drives 0 and 1 in the _same device_ (8, 9, 10, or 11), used in this format: ' "0:"..." ' or ' "1:"... '.
With the way most devs programmed for disk drives, either setup would normally be pretty useless, unfortunately. That sort of expansion was VERY underused for the 20 and forward, and that's too bad. It would be interesting, though!
"...Based off of this design..."
If they're based _off_ of that design then what design are they based *ON?*
I assembled this whole thing, but all I'm getting in the red light blinking three times. I'm not sure what is going on here. I've tried everything I can think of. Different SRAM chips, different CPU, different EPROM's.
I am in a similar situation; the serial bus hangs when I try to read it and the computer needs a reset to recover. I'm not sure what's wrong, but I'm waiting for a friend to finish his builds and produce a working device. Once that's done, we can compare part and swap components, etc...
@@SteveGuidi Same thing is happening to me. Serial bus hangs as well and the computer is frozen. Can't read any meaningful error messages. Turn the drive off, the computer unfreezes.
@@aa7aeThank you for the suggestion! I never considered that the SRAM might be a problem as it was purchased new from DigiKey. I'll have to test that out!
@@SteveGuidi I found out my SRAM is all good after all. My Retro Chip Test Pro was not working right because I had soldered in the wrong value resistor for the MOSFET. The self test kept failing which lead me to troubleshooting. Once that was corrected, now all my SRAM tests good. I thought I had a smoking gun, but now it would seem I'm back to square one again. I'm going to test all the logic IC's but I doubt any of them are bad. Whatever is causing this built not to work is quite a mystery at this point.
I got a message back from the board supplier in which he shared experiences where using modern equivalents of the logic IC, 7406 and 7407 may not work in some cases (U8, U11, and U12). E.g. Texas Instruments version commonly available on Digikey and Mouser. Problems seem to go away when using older NOS Fairchild variants. I'm am sourcing some of those on EBAY from U.S. Seller and will see if that makes a difference. The 1581 service manual describes the parts of 7406 and not 74LS06N. I would think they should be equivalent, but maybe the older IC's behave slightly different? We'll try and see if this solves the problem.
The 1581 is quite a powerful totem artifact in the 8 bit C= community. Good thing I have two of them. I'm not here to gloat, but I do have two 1581s. :) Nice project BTW.
The 90s *and 80 (since it came out in the 80s).
This is an expensive project to be sure, and I really have to wonder about how things have changed in the year since you made this video both in terms of making this project perhaps way more expensive and in terms of making an analogue potentially a lot cheaper.
What I mean is … 3.5" floppy mechanisms are still common enough, but I wonder if some of the chips you need for this kit are just no longer out there in quantity enough to make this viable. Something similar happened with the BYO VIC-20 kit. Made with all new parts … except the MOS 6560 of course. The VIC chip. At the time they were cheap and common enough on eBay for the intended use-replacement parts to bring a dead VIC-20 back to life. Not enough for a few hundred hobbyists to buy them all up hoping to build kits!
If you were designing a kit like that with the scarcity of those chips now (they sell for > $100!) maybe you'd start by assuming those parts are unobtanium and try to get one of the more powerful microcontrollers to emulate it. The RA4M1 (Arduino Uno R4) can handle 5v logic, but it's SMT soldering so maybe you want to just have pins for the Arduino to plug in. You're gonna get people calling it "cheating", but it's literally emulating one chip, so they can stick it. Or if you're willing to do a little more to interface with the 6502 bus, the RP2040 or STM32 are even cheaper options.
These are all SMT chips unless you get them with their common implementation boards, so that violates one of your rules of sticking to through-hole unless you make other more significant changes to the design. That said, if you're replacing a chip with a controller that can do more complex logic … can it also do some trivial glue functions without feeling like you're "cheating" too much? And is it worth "cheating" to bring the complexity and more importantly the cost down quite a bit without breaking backwards compatibility?
It doesn't look like anyone's done this kind of thing yet, but … if prices for things continue to climb, I imagine it's only a matter of time. Maybe that will truly make it a 1581 for all? Maybe.
Fortunately, the most expensive IC (MOS8520) in the trio of no-longer-manufactured parts is still around $20 on eBay. I bought a WD1770 a few weeks ago and it was a bit more expensive than when I first bought it, but nothing drastic (about an extra $2-$3). So it looks like prices are somewhat stable for this project.
That said, the Foenix FNX1591 is a new 1581 clone drive that uses an FPGA at its core along with many SMD components. I believe the only "old" chip needed is the WD1770, but there are plenty of those in supply. Not a kit and the price is close to what you would pay for a 1581 kit.
c256foenix.com/product/fnx1591/
I bet there are no components on the back of the board even if we _don't_ look there.
Haha... "via a via"? 😛
50 CAD (40 USD)?! I paid at least $250 if not more when i bought mine, but it was around 1984-85. Hm, your was not new ...
I delivered a lot of newspapers to earn that $50 😁. But it was indeed a used drive.
"2 inches in length"?
As opposed to 2 inches in diameter? Haha.