Hi, if you give me your email address I'll send the gerbers. It's not an "easy" build. I do have some spare PCBs but unless you're in EU it's probably going to be costing too much to ship one.

Thank you for your kind and quick answer! I just wrote you an Email.

@@sebastiankrosinger2101 amsat.org email no longer works, I need your address to contact you.

I have three different "vintage" programmers to be able to program all possible variations of TTL Proms (however I can suggest one rare and expensive programmer that can do them all). They aren't super rare, just maybe a big expensive (like euro 5/piece mean price). 2101/9101 are even cheaper nowadays. In my several-years-long experience, properly repaired boards last years (never got back anything so far).

@@iz8dwf I've just had a 50 year old 7414 fail after 5 years in service which I used it to fix one of my two pinball machines. What really puts me off besides the price is having to wait for parts and paying half the cost for shipping. I miss going to my "local" electronics shop (still a half-hour bike ride) and getting spares. Also, most of the machines I repair sit in the storage of my local arcade museum and I just fix them out of boredom (and because it's my favorite thing to do). My most vintage programmer is an 25Mhz 386 PC I pulled out of e-waste that just happened to have an EPROM programmer card. It can only do 2716-27010 (and 8748 MCUs for some reason), but at least it can go up to 25V which my other programmers can't. Also I can adjust voltages and speed.

@@senilyDeluxe Your programmer "sounds like" an old Hi-Lo (that I have too and I don't use anymore). I miss going to the local electronic store too, it is still in business but prices for old parts aren't much better than "ebay's" ones. Plus I have lots of old boards for parts, so in most cases I don't need to order replacements.

@@iz8dwf It's an ELV (which is a German electronics magazine which was also "big" in the kit business). I wouldn't be surprised if it was a rebadged model. The box with the ZIF socket is very small, I doubt it contains any extra hardware (apart from a red LED), everything is on an ISA card in said 386. I can't find any pictures, but I just found that ELV still has the firmware for free download (but it has a date of 1991, so it's probably the same version I have) (a couple months later, that e-waste container also had a UV EPROM erase lamp and an EPROM simulator which came in very handy for homebrewing on real hardware - just plug it in the board, plug the other side to an old laptop's parallel port, copy the binary and that's it - happy testing)

@@senilyDeluxe yeah at one point I was looking for an eprom simulator too, but in reality I've never so far needed one badly enough :)

The fact that there's no "smoking gun" IC and that two NOS ICs failed too (RAM and PROM) makes me think it was not about the history of this PCB. One really annoying thing about the italian clones of that era was that they used 74xx family liberally intermixed with 74LS and often in place of the much faster 74S. In my experience 74xx have a higher failure rate as they run hotter than 74LS.

@@iz8dwf I can't picture the root cause though. What sort of failure mode in the 7400 logic would cause this. Interesting to think about.

@@HeyBirtThe only plausible explanation I can think of about this particular failure is an abrupt change of the supply current on a large number of ICs due to the sudden loss of the clock. Trace inductances aren't happy about large current steps and they could have dumped enough energy to raise the supply voltage in some PCB areas enough to damage a few ICs. Other ICs driving shorted inputs failed shortly aftertwards. The amount of decoupling capacitors might have been enough to save some areas of the PCB but not others where the current step was particularly high. I can only speculate anyway. I regret I didn't take notice of the actual part numbers of the failed ICs, but I seem to remember most of the '157 were plain 74xx that are particularly happy to die if you stare at them too much.

well, I don't think so. It has no soldermask but that was a common thing in the '70s. Other than that it seems better than many other boards I've repaired in these years.

Oh really? Did you determine what triggered it?

@@iz8dwf My first assumption is that those arcade games suffered from high voltage surges applied with electronic lighters to gain free credit. That worked pretty well back in the days. But it can be a big ouchie if you're TTL chip.

@@_hackwell yeah but this board was working fine after just populating the missing parts. It failed while I was keeping it powered on as final check before packing it for shipping back to the owner. I failed to identify a single IC that could have started all the chained failure, that's why I lean towards a transient triggered by the crystal getting disconnected.

@@iz8dwf In your case I guess there was a power surge of some kind. Old TTLs on the blink die that way

@@_hackwell yes I agree about the power surge, but not originated by a power supply fault. That power supply worked flawlessly for more than two months while I repaired three boardsets.

Thank so much! It's always a pleasure to have a comment from the very one repairer that inspired me to start my own channel!

@@iz8dwf Thanks, you're too kind! You're an inspiration to me!!!, and many others too! Keep doing what you are doing =D

Indeed I love vector graphics too, I was exposed to Atari Asteroids too early in my life :D

@@iz8dwf Hahahaha. Me too !!!

Yep, used that a few times with old TVs :)

:D

Hi, assuming you mean pin 19 of UM3 (6522) and assuming you're trying to format a floppy with the N0 or N1 commands, the it means the CPU isn't able to address the 6522 correctly OR the 6522 is not working properly. If UM3 is socketed, you can try a different 6522 IC if you have one. R6522 and/or SY6522 ICs can still be found on ebay for example.

Tested two r6522, one new from Rockwell and another vic-20, with the same always high pin19 result. Today I will try another cpu. very grateful for the time

@@juanjoselopezgarcia4436 ok ,then unfortunately the problem isn't so easy. You should check pins 22,23,24,25 of the 6522 for activity, these are the control signals used to select and access the 6522 registers by the CPU (there're more pins involved though). Does the motor of the drive start when you issue the format command? Does the head seek to track 1 with the rattle sound?

@@iz8dwf Does the motor of the drive start when you issue the format command? Does the head seek to track 1 with the rattle sound? It does both things perfectly, this afternoon I will take samples of the pins suggested above.

@@juanjoselopezgarcia4436 if motor starts and head moves then the 6522 is addressed correctly, no need to check those pins. Are you sure you're issuing the correct format command and are you sure you don't have at least one short low pulse on pin 19? Old basic 2 command is OPEN15,8,15,"N0:DISK1,AA":CLOSE15 for example

however the italian clone has been named just "ripof" with one single F, so that was funny in my opinion

Hi Mark, thanks! Time is scarce, but I thought I'd edit some old footage eventually :)

Original 1540 ROMs are rare, you can try a search on ebay for example. If you really want to turn your drive into a 1540, you can burn the ROM image into a pin-compatible EPROM (MCM68766 for example). ROM dumps are here: www.zimmers.net/anonftp/pub/cbm/firmware/drives/new/1541/ (325303-01 is what you're looking for).

I really hope to find time (and space) to work on it. However I think AX.25 is extinct around my QTH since many years.

Yeah I witnessed a couple of aluminum electrolytics smoke live, but it's rare. Tantalums, on the other hand, do that a bit more often.

​@iz8dwf Correction... it wasn't the caps at all, or at least not directly. The 2SK1081 MOSFET blew up due to a short caused by the small leaking cap right next to it! Why the hell did PowerOne mount this next to a hot tranny?? Anyway, that part appears to be unobtainium now. Not sure why I heard two explosions, either...

@@NuGanjaTron Well, I did a big mistake to not inspect the power supply of my 2900 before power it on for the first time and I've been unusually lucky anyway. After this video I had to take the power supply apart again and substitute all the secondary side semiconductors and the voltage trimmer, seem the 2900 doesn't like the 15V supply drifting too much. The power mosfet is just 800V/7A part, I'm sure you can find a similar enough modern replacement.

@@iz8dwf Yeah, I now realise my mistake too. Funny thing is, the seller (classifieds, not eBay) showed it working and passing the self-test. So my guess is this thing has been sitting for years and was powered up once more for the ad, in the process of which the cap started oozing -- but not enuff to actually short the MOSFET. Anyway, the area around it is now covered in slimy gunk, so it must have continued leaking for some time after. Normally I check out PSUs before powering them up, but I really didn't feel like dismantling the 3900, and I now realise getting at the PSU is quite a pain. All in all, the design quality here isn't in the same league as Data I/O's older models (notably the 29B). Granted, the 2900/3900 are very capable devices and handle a lot of the more modern chips. Thanks for the hint on the MOSFET replacement, I'll check Alltransistor for that. I'll first have to figure out how to remove it from the heatsink tho; it appears to be epoxied (?) on. It also looks like one of the pins blew clean off! Incidentally, there appears to be some crusty tan residue on the outside of the heatsink that contacts the PSU frame; if it was thermal paste, it's _completely_ dried up.

Hello, maybe I can, but greatly depends where're you located. Send me an email message.

@@iz8dwf Hi can you give me your email

@@iz8dwf What is your email?

Can I have your WhatsApp number?

I don't know why my comment is deleted when talking about mail

Yeah, having common ground return inside PCB and to the power supply for both analog and digital circuit wasn't optimal, but using a differential audio path helped quite a bit. Atari used this differential audio "trick" up at least to Pole Position (what I'm servicing lately).

Hi, thanks for the comment! Indeed I didn't check the actual power, 3dB difference can be normal on these transceivers (as you said, lower voltage). Good luck with your repair!

Hi! It seems to me that not all stars "flicker". I've read somewhere that the starfield generator circuit in this bootlegs isn't exactly as the original one (which was implemented with a single Namco custom IC).

@@iz8dwf Thank you . Saves me fixing somthing that s not broken.

Hi, very good point. I haven't checked this aspect yet. Write gate to write data timings would need to be adapted to the actual erase head configuration if the drive hardware doesn't already adapt the timings. The standard floppy interface is "erase"-agnostic, so each drive must adapt the write gate delays to match the erase type. Of course the original 2031 firmware has its own delays to match the original head type used on those drives. I'll know more when I'll try using this drive with the new board.

@@iz8dwf What I really want to know is where you got that mechanism! Looked new in box.

@@8BitNaptime A german seller on ebay :) It was a lucky find indeed.

I fully agree!

For sure to me, vector-deflection CRT monitors have a special attraction. Maybe because I was exposed to Asteroids in my childhood :)

Some hints: preventive recapping is usually a bad move unless the capacitors are leaking (and yes, the low profile panasonic drives usually need new capacitors). 48 tpi disks must work fine on these drives, the HD ones do not work for sure. If you don't provide the ID, the disk firmware only erases the old directory and BAM (block allocation map) on track 39, if track 39 isn't formatted, then you have a "bad disk". If you provide an ID then the firmware makes a new format starting at track 1. It alternates between side 0 and side 1 on the dual side drives. You could use the old 10 OPEN15,8,15,"N0:LABEL,ID" 20 INPUT#15,E,E$,T,S 30 ?E,E$,T,S 40 CLOSE15 program to see what the error really is from disk drive firmware.

@@iz8dwf thank you very much! We managed to format and read back two 48TPI disks (HEADER "LABEL",D0,I69) by first demagnetizing them with a strong magnet, as you suggested in your video, using both the 8250LP and the SFD-1001! But the 8050 still gives WRITE PROTECT ON when trying to format even a demagnetized disk (and does not read the disks formatted using the 8250LP). The error reported is always "26, WRITE PROTECT ON,01,00,1". As you did in your video, we checked pins 16, 17 and 19: pin 19 never goes down, i.e. the drive never tries to write the disk. This sounds very strange!

@@DanieleCapursi for write protect error, check pin 22 of UK3

Woah... that's one of the most complex broadcast professional monitors :) I don't have one to check what might happen. But, looking at the service manual I see it's a grounded G1 plus cathode only drive, so blanking happens on the cathodes. Page 69 has the schematic for that part of the drive (I believe it's the neck PCB). It seems that blanking is done by Q713/Q714, when Q713 is on it should also drive all the IK-SW transistors and TRACE-SW transistors. No idea what might go wrong there without taking some measurements on the real circuit, it seems reasonably well made. Also since you see a "rainbow" it's probably an unfocused spot, so it might be that not all cathodes are conducting. Lots of unknown by just looking at a schematic really.

@@iz8dwf Seems like a fairly complex problem for my skill level. Nevertheless, I appreciate the explanation very much. It's cool to see things explained by someone knowledgeable, thank you. 🙂🙏

I'll try to make a better "SX64" internal view video one of these days if there's enough interest. PLAs do fail indeed, but only the MOS "compatible" ICs, the other SX-64 for example has the F-93459DC FPLA and I don't expect it to fail anytime. 6510 are reliable enough, I think I've seen two failed ones in all repairs from 1986 up until today.

Honestly I'd like to know as well

That was a cartdrige called "C64 BURN-IN TEST" that we copied in the '80s, no idea if it's available already for download (but I might copy it from my floppy disk. I also have no idea if the original cart had some diagnostic harness, we always used the program to check for non obvious faults after a repair (or confirm it was ok to return to a customer).

Grazie Marco!

L'unico componente che non necessita di adattatore e' appunto la PLA originale: 82S100, PLS100, F93459 e 556RT2 (clone sovietico della 82S100). - The only replacement that doesn't need any adapter is one of the original field-programmable PLA ICs: 82S100, PLS100, F93459 and the soviet clone of the 82S100, the 556RT2

@@iz8dwf tutti ormai introvabili (vergini)

Good to help Pete! The MOS part is a mask-programmed NMOS functionally equivalent to the 82S100, of course it isn't as reliable as the old ones. If you ever need an 82S100 replacement, I have a few spares left :)

@@iz8dwf Thanks Frank :)

Hi Mark! Sadly most of the new demos sound bad on the old 6581 SID. I do have a very good demo that was made for the 6581 but I didn't think about running it.

I find this topic very interesting since the '80s. But I always need to have a clean bench to play with my projects. So I first need to get rid of all the arcade stuff I have to fix :)

@@iz8dwf A problem I am all to familiar with :)

:D

I agree, it's a masterpiece :) If it has 4 x Z80 CPUs is a bootleg (original board has only three of them). You can also find lots of images on the web for both the original PCB and the clones. Most common clones are the ones like the boards I showed on this video.

Yeah that capacitor did quite some damage alone, but all the identical ones are gone, so it should not be in danger anymore. Probably those capacitors would fail more easily the more time they spent powered on and this particular drive didn't see much power in its life it seems.

LOL, it was surely a thing to be able to "see" how we could hack our drives in the '80s. And for me, the schematics were precious when people came in with defective 1541s :)

I have seen a raspberry pi (pi3?) solution that runs the 1541 firmware emulating the HW. 2031 and 1541 firmwares only differ on the interface part, so a similar solution could be emulating any other CBM drive (even the dual drives). However for me it's way more fun and interesting to build the "real thing" than to emulate it.