1050

https://www.jameco.com/z/68B21-Major-Brands-IC-68B21-PERIPHERAL-INTERFACE-ADAPTER-8-BIT-2MHz-DIP-40-pin_43414.html Back when I got mine they were Hitachi and top of the line while being quite a bit cheaper too. Watch out for bogus or bad chips on eBay and other discount outlets, it's a common scam these days. Jameco just won't go there. Do NOT select the suggested "you may also like" selections - they are NOT direct or any other kind of replacements. Only the 10 dollar chip is your huckleberry. Yes, these will replace all Atari 6520 PIAs. One of my 130XE came with the Hitachi 68B21 soldered in at the factory.

The yellow or yellow/greenish ones are capacitors and all of them can be removed or cut to enhance higher speed coms on the SIO port. This is commonly done and I would upon sight if it were my machine - you can wait until some toy doesn't work right of course and then do it, but I wouldn't bother fixing the lone cut off one at all. Blue ones are inductors, brown ones are resistors. Glass ones are bypass 0.1 uf capacitors with 104 markings. 74LS tells us nothing, it's the 158 number that describes the function. 74 is common for logic chip, LS is Large scale Schottky family describing the type mainly. One end of the resistor is connected to a green wire. Looks like pin 4 of the 158 dead bug mounted chip in photo one of post #12? Looks to be a good solder joint in that pic. Dead bug style mounted chips commonly don't have some connections made while other such as power and ground are soldered to the chip underneath. Dangling legs that go nowhere mean basically nothing. Start counting pin numbers from the notch end and left row all the way down that row, then keep count up as you go back down the opposite side. Pin one and the highest pin number will always be on the notched end. Nick Kenndy's SIO2PC moved site is here: http://pages.suddenlink.net/wa5bdu/sio2pc.htm This would allow you to use your PC as a source for internet files to write real floppies from or too. In the original form it used the printer port for the SIO2PC cable, but today much better to use USB2PC connections and perhaps PC software for that device. There are several to choose from if you are interested in such a system. Atarimax has one, http://www.atarimax.com/ and too many new ones to keep up with over here, but keep watching they'll get posted.

So MyDOS won't let you change parameters in that method. You can do it using other DOSes but just not MyDOS as it ignores the percom block by never asking for it in the first place and formats all disks by resending the percom block of the last entry in the drive tables for that drive number that can only be altered by the O Configure drive menu choice option before hand. I have no clue how you are going about the format itself for lack of information here, but if you were using MyDOS, you can expect to be ignored as to standard percom block rules forever. It's a bug and it's completely and totally broken like that and always has been at 4.50 and above versions. Might be that way for lower versions too, I haven't checked into it. I don't see any advantage to a half width 720 in 5.25 flavor myself, but I also don't have the interfaces that could/would use the one off drives including the 1.2 meg. As I remember someone had one those working fine on one of those early day interfaces and I don't even remember which one. While I bought two 8 inch drives for $40 thinking someday I'll know enough on how to use them, huh? Talk about a door stop, weighs more than two doors, need to get some Wheaties eaten first before you tangle with either one of those.

I only know what I'm told and hard info from you was a bit lacking, I'm not going to bring it up again while stating that most people have no clue how much more effective 90% ipa is when it comes to getting the crud off the head. Pray tell what is "that percom block get/set program"? Perhaps that's the issue if it were reloading default SD values after use? Could just as easily be the ATR operating system as well or the combination. Links to those would be appreciated. I can take a quick look at it, but a full on disassembly isn't a promise. Sort of an inside joke here, but you wouldn't be using MyDOS by any chance somehow?

Just alcohol ain't cutting it. Just like 180K isn't SD. I don't know what the issues are, but YOU are going to have to go to a store and actually put some coin down for a bottle that says 93% isopropyl alcohol on it before I'll ever consider you did it right. The juice that came with the disk was where you got ripped off, it just can't be trusted. 90K is SD, 130K is ED, and 180K is DD. If it says SD then something wrong with the report.

Likely indeed as it's a fact. Only the original very cut down first version of 95 that was only available as a 13 disk floppy set did not comply with the meme of making your own set of floppy disks as installation floppy disks from the distribution CD itself by using exact to the byte CAB files. And that's why windows install floppy sets were never for sale again. Win95 used 1,716,224 free bytes, 1,676Kb often called DMF formated disks. Win98 used 1,802,240 free bytes, 1,760Kb exactly sized CAB files for it's install CD, which required an 84 track capable floppy drive to make that particular DMF format with, toshiba ND-3561GR is one such beast. I can verify that they will produce the w98 DMF floppy as I have three. IIRC I made a set for 98se and it took 75 floppies to do it. WinME took it ever further, but by then floppy drive sales were already in the ditch and it wasn't a known or even mentioned thing to do anyway. I only learned of the ways to do it long after NT5 and higher were the only Windows supported. I had to find fdformat and fdread DOS executables on my own, but way back then it was still actually easy, I wouldn't hold my breath today waiting to find them quick. If you are cleaning a floppy drive head, use ONLY 90+% isopropyl alcohol, the lower grade stuff (rubbing alcohol) is actually a waste of time and effort. A good shelf in the store will often have both so peel back the peepers and find the real stuff.

Ah, shenanigans indeed. Confused only because those are recognizable as the US Doubler ram pack, but you don't need them since one of the big ICs on the top of the Lazer is 64Kb of static ram and the US Doubler ram is at only 256 bytes. I'm not aware of what happens when you try both ram at the same time, but I could believe a complete and total train wreak. But I never tried it myself. Oh my goodness, it's both upgrades in one and that's what the switch is for? Because you've got both ROMs too which is what this paragraph was going to be about until I remembered the switch booting up which ever one is desired. That would be interesting to see if it actually does that after the stepper replacement. Revision 7 is Happy talk, right there on the label. Ok, there is something you just don't see everyday...

None that jump out for me anyway, just check the voltages after.

Another story from the days of ancient electronics was just too darn much noise. They didn't filter it out because they were that cheap and the barely audible frequency they ran at meant the some people could hear the inductor's windings moving around. Frequency has been upped to 80,000 Hz or so and that makes the use of cheaper inductors possible while making filtering quite cheap and easy too. In short there are no such worries anymore, you will get pure silence instead. One sound that is missing today is the flash attachment on film cameras starting out at lower pitch and gaining in frequency until the voltage stored was high enough to trigger the xenon flash tube which also went off as a muffled snap that also was a unique signature noise. Way back when we could hear electronics working. And then along came better silicone with faster tricks it could do for cheaper. Your concerns and my stories are only out of date by some 50 years now, more recently I used to warn people that windows 98 sites were wrong in the first place and never updated in the second place. At least that doesn't seem to an issue at all today. I wouldn't say the 3086s were an actual issue from those days, it's just that I've been down this road before and new ones did fix that ONE problem for another guy. So there are a lot of really old 3086s out there still working great for everybody else other than you and that other guy. I did have to do some looking to find some and scored at radio shack for the handful I got a hold of. But then they used to have a really good eBay store before that folded up too. Maybe six months, they had everything a well stocked store would have had and I don't think they even exist anymore.

The ground you speak of is only the shield connection and it's only connected at the motherboard end, it does nothing else substantial than give the weak signals from the head during a read attempt a fighting chance to get to the motherboard without picking up some stray interference. No harm to the head at all. This is the regulator you might give some thought to. https://www.ebay.com/itm/255064973744 What I was going to do with it would be to remove the legs and mount them from the other side and this way "turned around", it would be "wired" correctly to drop right in with the correct in/out voltage pins lined up while still being moved away from the heat sink proper so that it can get airflow without taking the risk of the backside shorting out on the bare metal of the heat sink. These come in a 7812 flavor as well, but none of them seem to be as cheap as they were precovid (go figure). Only bulk purchase seems to be able to get these at close to $2 each mark as when I got them it was 79 cents with postage included. They were also very much easier to find since they were the only other choice then. Change out both 3086 and get some decent voltage on that board, my bet is this will solve the issues. Thinking about the damage at the jack and it's possible the previous owner tripped while carrying the drive in a disassembled state, rather than the brutal just yank on it method that might be assumed. Intermittent all gone now?

Well, I would order up some 3086 chips and a 7805 to change out what you got now. There are switch mode voltage regulators that operate much cooler than the 78xx versions do and those add a lot of heat to the wrap around heat sink. Now would be a good time to order some from china? Via eBay or any other distributor you like.

So hard to say at this time. Around here U7 is more commonly known as the RIOT chip, Sam's photofact people aren't Atari people so what can I say? When it goes bad typically (and they do) it's a nonfunctional unit to a completely fixed deal. So still zero on the intermittent symptom. U21 is responsible for creating the gain from the head on reading and also the bias and erase power to the head during format/recording IIRC. I would suggest a complete cleaning of the removable chips from sockets to carefully wire brush their legs, application of deoxit and reinsert into sockets at least 3 times to eliminate a cruddy socket from the issues it still could be. Shoot the drive switches with deoxit and work the devil out them while still wet, let dry and work the devil out them again as they will go bad right in the middle of command/data communications giving wild evidence that they are good when they are very dirty instead. Change both 3086 if you are going down that rabbit hole and hopefully the intermittent condition will resolve itself at some point giving an indication of where that issue actually is? And my bad on the intermittent issue, there was a case of dirty sockets just recently cured, I have now been able to recall. He posted quite a string before he did the deoxit thing with the sockets and bailed off the thread immediately when it started working again. It really can be that simple and no need to apologize for overlooking what are usually the first steps to do - they just wind up being the last steps instead and in life stuff happens. It happens to all of us too. Yes the 3086 is just five transistors, but back in the day they used kinda poopy silicone to make them with so any number of things can and do go wrong with them. Thompson/RCA/Harris were making them with the good stuff, ecg912 is their number for it. MC3346P, nte912, and ha1127 are other numbers used for the same 3086 chip.

Actually it does look exactly like some factory busy work for the fix it guys - there will be no book reading or food fights on my watch. http://www.atarimania.com/documents/atari-1050-field-service-manual.pdf http://www.atarimania.com/documents/Atari_1050_Disk_Drive_Sams_Computerfacts_Technical_Service.pdf Wish I could help more, see if you can grab a 5 volt sample if it ever comes back from the dead again? 4.85 might barely (3%) be within tolerance, but if it's better than that when it's working, I would change the Q7 7805 responsible - I might do that anyway. I can't recall an intermittent issue ever before this one and of course they are the worst to discover the reason behind it. Two 3086 chips U21 and U1 are known to get weak or quit outright. Will be watching for more of the story.

It's the binary switch so the copy executable can't treat any of them as a text file, once up front is all it takes and no take backs either. So doing it more is probably just ignored anyway. A proper batch file might be able to judge the size of the target file and assign the proper header to it making it capable of doing an entire folder of all sizes at one go. And batch when MS went with NT operating systems just became insanely even more powerful and useful. So much so I would have to lurk in those old forums for few days to even feel like I might have an idea of how to go about it now. And it looks like Russ Campbell did some of his QB magic and there is a solution now anyway.

So now I am confused again. How did you modify yours to suit theirs?

Fair enough, just about every "what is it?" I've seen in the last ten years has been a RAMBO, so yeah, you could say they were popular, while at that time I was unaware of any of it. I only came to the extended memory issue well past the introduction of the 130XE which set a new standard for extended memory use and that was the only one I had any possible knowledge of and for quite a while. As to support, I'm thinking it could be debatable. I do recall noticing in my early days of looking at ramdisk software that almost every ordered use of banking bytes was high (first) to low (last) values which bothered me when it came to the 4 banks of the stock 130XE ($EF, $EB, $E7, $E3) because that meant they were always the first to be scrambled beyond use if they held ramdisk data before because they sure didn't hold anything useful afterwards. So debatable in the sense that the only use I ever had for an extended memory sector copier would have been the likes of copymate 4 series and I didn't have extended memory then anyway. I got along just fine of course, watching a two pass copy and not really a care in the world about it too. I really was that late getting into the game, and what happened before that is dark matter to me. The point is - if it's still a high value to low value byte order that this software is using, then expanded memory could be used for the sector copy without using up the low value bytes which is what the RAMBO uses for the OS banks ($8F, $8B, $87, $83). By careful consideration of the RAMBO? Or just luck of the draw? I hesitate to speculate at all - it IS highly compelling that this high to low order was so very commonly found in my early days of investigating software and hardware and I did wonder about why it was so quite often. This would certainly help a lot in keeping a RAMBO machine viable. While I'm thinking the ramdisk didn't matter anyway since copymate was a reboot and anyone that left valuable information on a ramdisk expecting to use it after two reboots would instead have just saved the files to something more secure and also consider the ramdisk properly rooted afterwards as well. I don't recall ever even using the extended memory capabilities of the last versions of copymate after I got my newell one meg up and running in an 800XL. That would have been 1995, a year or two before my BBS days too. Years later yet for my first 130XE even. I had a long ways to go again before I got competent with machine language enough to tackle MyDOS source code and fix a couple of bugs in it with a MAC/65 cart. By the time I was actually up to speed, Atari was a thing of the frozen past. I never really got to see Atari as a live creature at all. It was around and I was around but our circles never crossed in the slightest. You on the other hand practically invented expanded memory on the 8 bit Atari, and Atari was so impressed with it's versatility, they made their own version of it. You were there at the beginning and the end of Atari as well. We can't be grateful enough of how you contribute around here, as you do it often when it's really needed too. Thank you Claus, it's always an honor and a pleasure to understand your ideas one on one like this. I have even more to think on and software to investigate...

Questions are welcome though. The RAMBO style in particular uses the extended memory for the original base memory to store 4 banks of the 64K of ram that the computer uses for the OS. So a RAMBO by default is never the full size of the memory chip, the usable extended memory is that minus the 64K of the system memory so most copier programs will just trash your computer if they are tried. Very little support was given the RAMBO type of memory upgrade and when it's Claus's original 32K bank version instead of the "standard" 16K sized banks, you can be assured that nothing currently available can work with it - as is. Outside of the driver Claus supplied in the article there is nothing else that works with it. A standard 16K Rambo type of memory upgrade has little to no support either. We have just recently accounted for the RAMBO type of these memory upgrades in memory testing programs for example and nothing else to my knowledge. So a RAMBO sector copier then would not exist as such. Painful to understand this is what you have found since you appear very keen to use it. But a better upgrade would give you quite an advantage as to actually using an upgrade that does work with existing "one go" disk and sector copiers. Or my favorite use as a ramdisk. The wiring used on the backside for "circuitry" is identical to the Williams cartridge wire (and just not seen in the supply chain today) so it leads me to thinking that who ever this Williams guy was he was certainly a player behind the scenes somewhere. Almost determined to be a ghost too. I would consider that upgrade to be a collectable only and it's value ruined by trying to make it into a standard 16K banking RAMBO. Please start over with a memory upgrade attempt and use something other than a RAMBO basic design so you get back the use of the stock supplied memory banks for OS use and thus compatibility with currently available software too. One example is the Newell 256K upgrade: So H1 is a wired socket that replaces a current in use 158 and H2 was supposed to plug into Newells mind I suppose. H1 makes sense, H2 does not. Atari certainly wasn't going to supply a motherboard with an H2 socket but they sure could have if there were bound and determined to make Newell a billionaire? There are other schematics of other 256K sized upgrades, but none are falling off a log easier. This one was just well known. At any rate, best of luck whatever you decide to do.

Found this, can't vouch for it as I've never seen it before and believe me I've looked before and a lot. http://www.dosdays.co.uk/topics/floppy_drives.php But sizes aren't in the info. You'll have to see first hand.

It's not broken, you are just trying to use it wrong. In order to go to DOS, you must first KILL the wedge and it's just the nature of the beast. If your wish is to go back and forth often between BASIC and DOS then you might need to write a .BAS file that will load and invoke autorun.sys file again after you have entered BASIC by doing a DOS Run at address command (to get you back in BASIC) and then do Run"d1:startme.bas" The way it works is a bit off, going to DOS should first KILL the wedge automagically. And getting back into BASIC with the wedge active again shouldn't be a typing contest either.

Hi Larry. No I can't explain the double colon deal unless 4.53 isn't clearing the line and this is actually the D1:MISC as the default directory, when it used to be a bunch of others and we are looking at what used to be there because none of the tail end is being cleared. I seem to remember an issue with the drive report line (one line above) where these kinds of mistakes were going on, misplaced resulting in no spaces, wrong data, not cleared, etc., and that was the part that got me to take a good close look at 4.53/4 to notice that DOS.SYS overwrites itself with itself at an offset. And I couldn't make the source do that or anything for that matter. I had to call it a crazy making waste of time to work on that one. They are corrupted files. David's ideas in 4.53/4 are just brilliant so I just gave him credit in the source and stole (and fixed) everything he offered to put it into a version of 4.50 that I had already placed several bug fixes into which today is 4.55 beta4 available at Mathy's MyDOS page. https://www.mathyvannisselroy.nl/mydos.htm Just like you, I tried everything in the book to wind up with colon(s) after, and it's not possible. Let alone a repeat, so I'm stumped big time. But not enough to boot 4.53.

SDX is a cartridge so it won't be an ATR unless it's sole purpose is to program an AtariMax flashmax cartridge (or similar) into a SDX cartridge. Disk based SpartaDOS does not have MyDOS subdirectory access. http://sdx.atari8.info/ Is where you can find SDX files and information. In your first post, the directory MISC is on drive 2 so doing R D2:MISC should work a lot better. You can enter * instead of a drive number to see the directory list of the files in the default directory as well. Charlie Chaplin is very much correct in his second paragraph of post #6, in order to have displayed as your default directory of D1:MISC::MISC2 something of an egregious nature has happened to that boot disk. And I can't imagine what as that particular situation should be impossible. Unless 4.53 has a simple display issue and that's the only mistake. But it does appear that you have an excess of subdirectories named MISC just about everywhere. That can lead to many confusions if you haven't yet noticed.

Because it's using page 6 in real time to do what it does. You'll have to get tricky and assemble the output to a disk file instead and also use some other debugger to debug with as that is also likely to be in use if any OSS cart is in use as well. Just a WAG for the second caution however as I don't have a great deal of experience with OSS carts, but they tend to keep to a habit if you will. Directions will be found to the contrary which then have to be disregarded as lies unfortunately. Such was the case when I forged ahead and needed to construct a pre-swap file that must be loaded first when I want to hot swap my mac65 cart in when I didn't boot with it in the first place. My hardware is modified to allow that with a switch thrown. Load the file with page 6 set up as it is with this cart boot and the hot swapped cart doesn't know the difference. Try the hot swap without the file first and you get garbage and a locked up machine. Therefore it IS using page 6, so you are not allowed to overwrite page 6 without dire consequences to the behavior of that particular cartridge. Writing code directly to memory is the default mode with mac/65, it can be changed with alterations to the ASM command as found in the manual.

https://www.jameco.com/z/68B21-Major-Brands-IC-68B21-PERIPHERAL-INTERFACE-ADAPTER-8-BIT-2MHz-DIP-40-pin_43414.html This is an equivalent and suitable for direct replacement for our 6520 PIA. eBay might be loaded with fake offerings which is on the rise over there, so very glad I purchased full tubes of these and others before scammers had a reason to exist in that trade. A is purportedly a 1 MHz design and not good enough for our 1.79 MHz clock. B is purportedly for 2 MHz and is supposed to work, but I won't stand behind it. I'll stand behind it when they SAY it's 2 MHz and only then like when Jameco says it. And I trust they are NOT selling fake chips too. P usually stands for Plastic (instead of ceramic) case so who knows given there are at least two flavors available?

So oddly enough 120ns is perfect for a game cartridge but put it in the OS slot and they become quite flaky and often won't boot at all. I can't claim to know why either, just what I've experienced and I see Steve at AtariMax using pretty fast flashroms which I also can't get working in either spot for my own projects. Would really like to know why it is that way. TBH I've done just enough to know I don't have much luck with fast ones and I ran out of steam at that point, if it were more important to me maybe I'd still be working on it. But I'm stalled and not very active any more, so not interested enough to make any of my own progress on that why question. It's more of a curiosity to me nowdays.