at 10:58, I have to say this is the BEST (and only?) illustration that I have ever seen on TH-cam of how a floppy drive head writes data to the physical disk. VERY well done! I'm creating a video that will cover an overlapping topic, and you have inspired me greatly here! Thank you!!!
This is not just a video, is a tech document. Your work is amazing, Noel !!!!! I trick I learned time ago to press that metal piece in position is inserting a small one twisted 90 degrees inside the same hole, between the piece itself and the motor structure (the round metal piece), in a way they make a "cross" and one pushes the other. Then, try to keep them together with glue or solder. It worked "some years" ago....
Brings back memories, back in 1984 I was repairing FDD at Mitsubishi on 8" floppy drives, then 5.25" and then 3.5". Radial and azimuth alignment was the norm. The key was getting the screws tight as shipping (trucks bouncing) could easily misalign the heads, specially on the 3.5", you need to start tightening the screws when the head was out of alignment, and tighten into alignment, most techs just could not get this right. Adjusting rotation speed, track 0, disk clamping force were all adjusted. I miss the Brikon FDD tester analyzer. www.techtravels.org/wp-content/uploads/2008/12/brikon723-1000x1.jpg Much more fun for me was repairing the circuit board to component level, thats why I moved to PC motherboards from the FDD department.
Good tutorial. I am currently restoring a completely busted MSX2 (Yamaha YIS-805). Both disk drives have broken/missing diskette presence and write protection sensors and have a lot of rust and oxides inside. I had to completely disassemble the entire disk drive and clean its parts from rust. After assembling back, the drive obviously needs re-adjustment. Unlike the drive you showed, the JU-253 drive has track 0 sensor that needs quite a lot of precision to adjust correctly. But it's doable even without the oscilloscope. What really helps is to mark the exact position of the parts before disassembling with either a marker or with a sharp object. This saves quite some time because after you align everything as before using the marks, only very minor adjustments are needed to make the drive work again.
Oh this one has a really fiddly track 0 sensor too! It's just that it was in the right place and didn't have to be adjusted. If both the track 0 and the motor alignment are off, it becomes quite challening. That will be a topic for another video I'm sure.
Great job! To help rotating the motor more precisely, and also have the screws more thightened while rotating (so the motor does not move when final thightening), you could 3D print a lever, and clamp it on the motor. This way, you have more torque and more precision. I will try to design something this week-end.
Excellent video Noel! I have recently subscribed to your channel. I am currently restoring a CPC6128 as it was the first machine I owned as a child. I really appreciate you sharing your outstanding knowledge to help me and others fix up these old machines. Keep up the good work!
I did the same thing, but on a 3.5" double-sided drive. I didn't use a computer in the process at all, instead I used another drive, to record a square wave pattern on both sides of dist at track 0 (using function generator), and a silence on track 1 (actually you can record even analog signals on the Shugart floppy drive). Then I used this diskette, to align track 0 and top head on the drive being repaired, using the oscilloscope.
@@NoelsRetroLab I don't know the Amstrad floppy interface, but believe it doesn't differ from a Shugart/IBM. If so, then you can control the drive with a simple Arduino sketch, or even with some switches for direction and side select, and pushbuttons for a step signal and a write gate.
I feel like a fixture that could snap around the motor, containing a spring that could sit on the back of it, pushing in on that little tab would be an easier fix, and could prevent it from happening again... cool vid!
I like that. It would have to be pretty thin to fit in the case, but it seems definitely more robust than bending the tab like I did or using some kind of tape like some people are suggesting.
@@NoelsRetroLab I think, the flimsy tab won't survive re-bending very often anyway. It seems like the motion of head assembly deforms the tab over time because it pushes it back. A bit more support for the tab should solve this problem. Look at the other drive model with the bigger backplate on the motor. How is it solved there? Great video!
While watching the vid I was thinking the same: a metal strip ~2cm wide, shaped like back of that motor, and held by the same two screws, and either a spring like Scott suggests or even part of that piece bent to reinforce the tab. I wonder if it can be improvised bending some old Meccano.
... that was my first thought, too. At 10:37 you can see the washers still in place. But I think as an educational video its fine to explain the alignment preocedure in case you have to exchange the head- motor.
There really isn't any need to even remove the motor. Just slide a wire behind the tab and bend the font in as you hold it out. Then release and it should spring back forward again. Same issue happens to some 3.5" floppys for PC's. If you want to remove motor, mark the position.
You have no idea how much you helped me out with this video. I am a noob when it comes to computers, but I have an old 386SX laptop with a broken disc drive. It turned out that it was just slightly seized due to never been used. but I had no idea about this alignment issue with floppy drives, and needless to say my drive doesn't read anything now (I removed the stepper motor to grease things). My issue is, it is a laptop, so I can't just simply remove the drive and fiddle with it. And disassembling and putting it back together like a hundred times after every adjustment is just not feasible. So what I am thinking is getting an extension cable, and route the drive outside of the laptop, and do those adjustments there. The only cable though that I could find is an extension cable +board combo for Raspberry Pi. Also the cable has two female ends, so will need to do some soldering for a male connector. I will try my luck, as I don't see any other options.
thank you for posting. before disassembling the motor to bend that little metal tab, i might have tried just pressing on a postage-stamp-sized square of aluminum foil tape. if that little tab bent once, it'll probly bend again. al foil tape is available from home center or hardware store. good n useful instructions for stepper motor alignment. never tried that. i wonder what the little teeth around the base of the stepper motor are for. perhaps to resist rotation of the stepper motor housing ? please take care n stay safe.
I considered trying to tape it in somehow, but I was afraid that would come loose sooner than the bent tab itself, so I opted for this solution (and got to align the motor at the same time). But I'm sure that something could be done along those lines.
I don’t think I’ve ever seen a 3 inch floppy drive before, so never worked on this type. I was a service tech for a big computer dealer back between 1979 and 1985 or so and did work with a lot of 5.25 inch and 3.5 inch drives though. We used to realign a lot of the Apple (Shugart) 5.25 inch drives and later some of the IBM full height 5.25 inch floppy drives (Shugart and Tandem). For the 5.25 inch drives there was a special alignment diskette available from Dysan that we used. It was actually recorded with the test tracks off center, so on a scope it showed what was known as a Cat’s Eye pattern. Basically the pattern looked like two different sized ellipses on the scope, and you adjusted the motor until the ellipses were equal in height. Your methods seem to work well enough though, you could probably even do that with a digital voltmeter set to AC instead of a scope, as long as it’s input impedance was high enough that it didn’t load down the signal. I realize why you went through the process as a teaching experience. For my own equipment though, I think I would’ve just glued something onto the back of the motor to hold in on the lever and avoided the realignment? Thanks for the video.
Cool. Someone else mentioned a special disk. Even a disk with all 1s in one track and 0s in another track would be very useful zeroing in to the correct alignment. I might try that next time. You're right that there was probably something external that could do the job. I wasn't sure what would be permanent enogh (or at least as permanent and bending the dumb tab, which is not very), but it was fine since I also wanted to show the alignment procedure.
My memory is a bit fuzzy on this. I started out using the analog cats eye method and then switched over to a digital alignment disk. It was faster because you didn’t need to hook up the ‘scope. Don’t remember how the digital alignment disk was actually used. Couldn't find a source for the digital disk and only one New Old Stock analog disk on EBay.
I only used the analog diskette so not really sure how the digital worked. I haven’t looked through these yet, but got several hits just googling “Dylan digital drive alignment diskette”.
This brings back some memories of me fixing my amiga floppy after it stopped reading disc. Unfortunately i don't have a oscilloscope. But i got the discs to read. There were to brands one was sony and the other was chinon. Thanks for the video
Some great info here. I never knew the motor just had a flimsy piece of metal to keep the spindle aligned. Bit of a trick, btw... You can actually hook up an audio amplifier to the signal from the read head, and "hear" the tracks. It's not a very scientific method, of course, but if you don't have an oscilloscope, it can help you get aligned with the first track. You can even doing this by hooking up to the /READ signal on the ribbon connector (instead of directly to the read head, where the signal voltage is tiny). But, the /READ signal is +5V logic level, so if you hook up to an audio amp, definitely add a series resistor (or around 1K) and a small capacitor (say 22uF) for DC blocking.
@@NoelsRetroLab Yeah, that pretty much sums up the Amstrad build quality back then. lol But I still loved my 6128 and Speccy +2 in the 80s/90s. ;) I bought a 6128 a few years back which had a faulty drive. I replaced it with a Gotek, but it might be interesting to see if it has the same issue. Although I no longer have any 3" disks to try, just the one that came with it.
I have 2 suggestions : First you can draw a mark on the motor before disassembling it, then you can put it back at the same position it was before.. Secondly, please use silicon based oil (or grease) on plastic.. Some plastic are "dissolved" by petroleum based products.
On IBM drives at least I always just made a reference mark on the stepper, then marked the low and high point where the drive would start to have issues and then locked it down in the middle of those two ranges. An oscilloscope of course is the best way.
Hi Noel greetings from Australia! Really enjoy your videos especially the troubleshooting ones. I'm an electrician by trade and really one day want to get into electronics myself, but to a dumb wire pulling electrician - electronics can seem like pure magic at times - especially vintage analogue circuits! One day your vids will push me into buying an oscilloscope 😂 Keep up the good work my man! I think you are doing a great job 👍
Thank you so much! It's not as complicated as it sounds, and you have most of the basics down! And honestly, vintage electronics are way easier than modern ones. Much easier to understand and diagnose. I didn't know real, practical electronics to repair computers 5 years ago (I came more from the academic side and the low-level programming side), so you can definitely pick it up and get started! Good luck with that!
I've never done it with a 3.5 drive. Is it a similar process to realign a 3.5 disk drive, or is the mechanism very different? I imagine they still have a stepper motor and a worm shaft at their core, right?
Excellent video! I see a few people have mentioned fabricating some device to keep that motor tab pushed in, here is my suggestion. Find a non-functional audio cassette, remove and modify the metal bar and felt pad assembly that maintains pressure on the tape head. Cut down the felt pad to the right shape to sufficiently push in the tab . A slight lengthwise “V” may need to be bent in the metal bar to stop it from flexing. Solder the bar in place. I think the metal is copper or brass ... will that solder to the motor casing? A tiny bit of epoxy on the modified felt pad may be necessary to stop it fraying (dry completely before installation). That tab design is a really bad example of design cost reduction.
Hello, this is a great video, thank you for posting it, it is very instructive. I left a mark with a marker pen to keep the position of the stator. and I was successfully able to CAT the dscs even after removing the stator, and that is still possible, as you mention, in several places. I have 2 questions 1) Does moving the ring that is screwed on the other side of bushing have any influence on the indexing? 2) My motor is only moving outwards, it cannot / will not move inwards, and I do not understand why as I "rewet" the solder positions of cable of the stator. I cleaned the shaft and the stator and it all rotates OK. I am inkling towards a problem between the stator and shaft but I am puzzled. Thanks in advance for any advice you may have. (FDD is an EME-155)
Nice video, I'm working on a floppy project at the moment, so this was a nice video to help me understand the stepper motor a bit more and now I know what the T1 and T2 points are!
@@NoelsRetroLab Im trying to build a replica of the Amstrad SI-1 that was supposed to have been released for the ZX Spectrum +2A but never was, to add floppy support. I would love to use 3" but they are difficult things to come by, so I'm learning how to adapt an old 3.5" PC drive - really got into my Shugart spec and READY Vs DISK CHANGE signals! My wife thinks I've gone mad! Also though noone had made a SI-1 before, but then I found them on sale at 8bit 4ever (I think that's the name of the shop) and it looks really good, so maybe I'll buy one of those of it doesn't work out. Was soo jealous of your broken 3" drawer 🤣 so many drives... 🤤
@@mogwaay A lot of people just use 3.5" pc drives on the speccy +2A/3 as drive b, I used an old amiga external drive, I took the internals out followed a guide on making a cable with the ready signal, I also followed a guide on making the drive switchable from b to a so it can boot games and I cut it down to the size of the speccy +3 so it can fit on top of my +3 drive, I painted it and put a +3 sticker on, looks really nice, here is a link if you are in the SFEO group :-) m.facebook.com/groups/spectrumforeveryone/permalink/927593487587123/
@@Digital-Vortex Hi Chris, thanks for that. I've joined the Facebook group, not heard of it before, so thanks for the tip. I finished wiring up my prototype SI-1 clone and it seems to work - my +2A thought it was a +3 and said drive A: was available - woo hoo! At the moment I'm just separating and grounding the ready line at the moment from the drive and ignoring Disk Change from the IBM 3.5" drive and the Speccy seems ok with that. Unfortunately, at first attempt (at part-my-bedtime o'clock!) the floppy drive wasn't reading the disk (couldn't find track 0) or letting me format, but it did seem to be trying, so I'm taking that as a win for now and will look into it when I get next chance. Cheers!
My new favourite TH-cam channel. I've been reading your Amstrad diagnostics tools on GitHub. What a contribution!! I'd love to know however where can I purchase the dandanator?
I am surprised they don't have a self alignment feature. I am used to steppers on CNC and laser cutters , and the "home" when they start up . great bit of info , thanks
Good point. I guess in order to self-align the motor needs some way to rotate in small increments. I suspect that's not possible to do from software (otherwise you could run some sofware that would mess up your drive), but it would be very interesting. I don't know if it's a limitation of the controller or the stepper motor itself. If it was just the disk controller, I could imagine trying to write a self-alignment program with an Arduino. Hmmm...
@@outbackchillin5628 Here too, there is a home sensor: the optical one, on track 0. But the resolution of the stepper is not enough to move between tracks, perfect alignement is needed between the steps of the motor and the tracks position.
@@NoelsRetroLab I think these old small motors only have a few full steps per turn (modern hybrid steppers has 200 or 400). Modern steppers also allow drivers to do micro-stepping, by not energizing coils at zero/full current, but in between, so the stepper does partial full step rotation. On the other hand, old stepper motors don't allow micro-stepping, and only turn half steps. This is old technology, so nothing much to do.
Another great video Noel! I don't have a 3" disk drive or even a CPC, but I enjoyed it anyway. I agree, that design is ridiculous. Someone was pinching too many pennies. Also, happy channeliversary! I didn't realize you've only been doing videos for one year. They are very well organized and made. I look forward to the next one and many more!
Thanks! Yeah, typical Sugar I'm afraid. Always looking at the bottom line. I'm actually amazed that the Amstrad CPC turned out the way it did when that was the driving mentality at the company. And thanks for the anniversary wishes! Crazy! It seems like a long time and not very long at the same time 😃
It's odd that the outside diameter of the motor has gear teeth, maybe there was a tool like a screw driver with a small gear on the end and a pin to fit in a hole on the motor mount. I have done the same tweaking you did to fix the motor position but 30 years ago, but on the ancient Tandon Tm100 5.25 inch drives. CLUNK CLUNK BZZZZZZ lol. Very old but fun to fix, and they had a strobe disk on the disk hub wheel so the speed can be set. The plus D disk drive interface for the speccy worked on old and new drives.
What a great explanation of how to check/repair the 3" drive. I imagine that it would be the same for a ZX Spectrum +3. So, my question is, is there also some test to use on the Dandanator for ZX-Spectrum, like the one u use for the Amstrad. Thx :)
Test for moving the head for the disk? I'm not aware of any, but maybe. In general the Spectrum didn't get nearly as much disk support as the Amstrad did.
I don't have any special knowledge or anything, but one thing did occur to me about about aligning without an oscilloscope - you won't know if you've actually aligned to the centre of the track. So perhaps it could work on most disks, but then you might write to a disk and have it not read correctly on another drive, or find some disks don't work and assume they're bad, because those drives or disks had their own slight misalignment which might have been within tolerance for a correctly aligned drive but not for your slightly misaligned one. Not sure. I'd have thought using a scope would be best if you have one.
I would have tried a relay adjustment tool to bend the tab in before removing the motor. The tool is a slim rod or bar with a slot on the end. Put the slot over the tab at the attached end and twist to force the loose end against the motor shaft. If it didn't solve the problem you still have the option to remove the motor.
Interesting video. I have the same issue with a 128 setup I got given last year to add to my collection. I'll try the metal strip and if that fails, the motor as I've already changed the belt. It's like aligning old radios which I don't tend to do as I don't have an oscilloscope but the slower method, I'll have a go at. Thanks ps - subbed.
Thanks, glad it was useful. One thing to watch out for: Do you get Drive A: before the error when you type CAT? If you don't, then it's likely the track 0 sensor. If you do, then it could be an alignment problem (or a head problem, or a data communication problem). So always start there.
When I did something like this decades ago, I thought it was important to align it with track zero starting from the edge where there is no signal at all, otherwise you could accidentally align it to an adjacent track making every track one off the correct track. Perhaps the software is verifying that it's reading the correct track?
@@NoelsRetroLab By the way, it occurred to me after i had posted my comment that it might be helpful to mark the position of the stepper motor with a fine-tipped sharpie before removing it - i.e. a mark that covers both the motor and the chassis; that should make it easier to realign it afterwards, don't you think?
@@BertGrink Yes, I think it could be helpful as a starting point, but it's so fine that even that will need to be adjusted. But it seems like a good thing to do anyway. Next time I'll try that.
Yeah totally agree, would've loved to know this stuff when I was a teen, but alas it was just so hard to get information like this... It's great to come back to it now, have been really inspired by Noels videos to give it a go.
Noel, if I was to detail the full procedure to you and how to set up TS00 after adjusting the Stepper, would you be interested in making a new video? I’m sure you could use this video and cut in new footage. Also checking the index pulses at 200ms. Personally I think this would give the full story for 3” drives and how to set them up post service with lub and belt.
Nice. Since you had a partially working drive, would it have been possible to measure the voltage before removing the motor and then you would have known roughly what you were aiming for when you reattached?
Good idea. I should have done that! (or measure it now). I'll probably do another video on these drives at some point about adjusting the track 0 detection, so maybe I can do a quick followup addressing things like this. Thanks.
Great job! I would have really just took a chunk of plastic and fixed it so that tab was solid against the back by gluing the plastic to the back or something like that. lol
Hi, I have a 6128 FDD that sometimes fail to read discs, and sometimes also randomly erase CAT tracks when I just want to erase one file. I'm going to apply your adviced to try and fix it (from this video and the first one too). But my drive also have an issue that I don't think you've covered in your videos. When I push the eject button, the drive sometimes fail to push it out, and I have a hard time to remove it. I guess maybe a spring is to weak to do the job ?
Thanks for another great video, I like the way you do the graphics explaining what's going on with the sectors/tracks :-) quick question though, Is it possible to use a multimeter instead of doing it by trial and error or oscilloscope? Also is there any software to help aligning for the spectrum +3? Thanks :-)
Thanks! I'm not sure you could use a multimeter, although Mr. Guru in a comment above is explaining that you can, at least on Amigas, so check out that comment if he writes back. I'm not aware of any software for that on the Spectrum, but there has to be some. If not, you can always move the drive to an Amstrad, do that, and move it back 😃
Hi! Thank you for this video it really helped understand a lot of things. I have a question regarding the test room. I purchased a M4 board from Duke and wanted to know if I can load a test room there and where to download the rom from?? Thank you again!
Good job!!I I just received an amstrad cpc 6128 today, and when power on the floppy disk is always with the led turned on. Do you think my problem is the same?
Ah that was an interesting video Noel ! and it's given me some ideas too, when i eventually get all my source code for Disced finished I'll produce a version that can be used for just setting track/sector(s), to help with testing this type of adjustment, that would be cool ! thanks for yet another great video :-)
Actually, someone in another comment had a great point: The oscilloscope method would be a LOT better if the drive were constantly seeking to a particular track (instead of having to tell it move forward, move backwards). That way you can see your changes in almost real time. So there's your unique angle for that software. I would totally use it!
If the drives works well when pushing the tab on the back of the step motor, what about marking precisely the position of the step motor before removing it. There are many "teeth" on the step motor, so just by putting a dot on one of them and tracing a thin line on the chassi corresponding with the choosen teeth tip, it would be possible to put it back just in the right position? Of course this could only work if the drive reads correctly when the tab is pushed, or else it means the motor has been removed before and not align correctly when put back in place. But if this is the case or you just forgot to mark the step motor alignment on the chassi, this method is very useful and works great! :-]
Thanks! If we're talking about 3" drives, the other models have similar issues, different solutions. The motor is held in place with a kind of rubber ring which tends to become much after many years of use. No tab to bend in there to fix it, so you need to get creative. Fortunately those drives are sturdier and that doesn't happen much.
i wonder if they make really small c clamps with rubber pad on the screwy end to just clamp around the motor or 3d print a clamp from polyethylene to push on thecasting flat and go around and push the shaft spring
That's exactly what the earlier models have inside the motor casing! Eventually they degrade and become a mush and the drive goes out of alignment too.
Realizing that putting it back into alignment was the focus, since all you did was push the tab in, was there a reason you couldn't just push the tab in while it was still attached, if you didn't want to go through the process of realignment? Or is there also a technical/electrical reason you didn't want to do that?
Good question. I didn't explain that very well. Here's the situation: That's a springy metal tab and it's 1mm away from the motor. You can push it all you want, but you won't be able to bend it in so that it's rest position is past the motor (so it pushes slightly against it). See what I mean? You really need to push it past that point and then have it back into place.
@@NoelsRetroLab Ah ha. No, that makes sense, I just had a mental model of it that was slightly off of reality. I wasn't taking into account the inner cylinder (at 10:20). When you said it was the magnet that was rotated I didn't think it had a direct impact on the tab. Looking at it again, with your comment in mind, now I have a better idea of what the inside looks like. Thanks for the reply! :)
I found an old drive that the head assembly is loose from the rail, so the motor spins but the head barely moves. You can move it back and forth with your fingers without the rail moving at all. I'm not sure if there's any way to fix that?
The advantage of the oscilloscope would be to monitor the waveform in realtime as you rotate the motor. Without having to restart the read every time. I'm not familiar with the CPC, but I expect there's a way to command the drive to read continuously, or even if that's not possible, just write a simple loop which reads a certain track over and over.
@@NoelsRetroLab and use an analog (needle) voltmeter, a sensitive one like the ones used for aligning vintage radios..watching analog needle movement is perfect when searching for peeks..much better than scope or dmm
@@NoelsRetroLab But how the signal can be refreshed in real time (as we can see jiggling on the oscilloscope), but not been really refreshed regarding to the amplitude? Is there some sort of buffer, holding it? I'm puzzled...
Hi, could you do a video of adjusting the azimuth on the Amstrad CPC 464 tape deck please? I managed to record an azimuth adjustment program to tape and it helped slightly but games still fail to load, I'm looking for an accurate way to adjust the azimuth. Thank you.
thats interesting.. i remember changing the head in my 464 back in the day and never needed to align it in any special way.. but maybe i just got lucky! i also 'made' one of them tapes with the head in it and a stereo cable comming out.. and never ran into alignment problems with it playing from a stereo system.. (dont ask me why i made it or was using it that way...i was just screwing around!)
good video, I was thinking, if that failure point on the stepper motor has got lose with time, it might not be better to mod it? you could easily weld a metal strip with a spring, to keep it always tight
@@danielhaupt2066 yeah true.. i forgot some ppl use weld and solder interchangably! :P ..needs a plate like the other models.. im guessing they have a spring to push against the shaft..
excellent content Noel!.. as usual!.. ..that stepper looks soo damn cheap! (more sweet Sugar goodness!) i was amazed the windings come away from the shaft like that..but i guess its not that bad of a design considering they had to align the head somehow...
Hi, at 4:26 you speak abouth drive models on the CPC forum. I've searched like hell and didn't find that posts. Is it possible to place a link to that specifique page please. Thanks :)
what about 3d printing a cup to cover the back of the stepper motor and let the cup press le metal leaf? the cup could have a screw to press fit the cup to the motor and maybe a small dent at the center to better press the metal leaf
I once realigned a 720K 3" drive from a PCW9512, using the trial-and-error method. It's time consuming, particularly because the adjustment of the motor is so sensitive - you have to apply pressure to it, but if you feel it move, you've gone too far.
I considered that, but how would would a tape have lasted putting the right pressure? I thought it would be better to do it right. But maybe that would have been enough.
@@fft2020 Yes, very possible. Add some rubber bit to push the tab in and the solder a metal piece to the casing holding everything in place. That would probably work.
@@NoelsRetroLab yes, it looked to me like the rear of the motor on the later version that you briefly showed us did in fact have some kind of plate affixed to it’s rear instead of that tab. As you say; dreadful design for that bit.
Thank you for the great vid, would it help to mark the motor and the chassis of the drive with a marker pen 1st before taking it off 🤔 and I have a few drives here but 1 has a white plastic front and 2 heads, 1 on top 🤔 does anyone know what machine it’s off. Thanks.
That tab is maintaining pressure on the motor drive shaft. It looks like it's not quite up to the job. I would add something to the back of the motor to assist the tab. Otherwise the tab will probably go slack again over time? But, this video is far more informative than just a bodge fix.
Yeah, I suspect it'll eventually go out of alignment again. Not sure how long it'll take though. Next time I'll try with a rubber ring on the inside, like earlier models.
7:15 i am wondering why to realign anything in the first place. If you have track 0 detection this means to me that drive at boot doesn't know where it's at, so it backs step at the time until it gets confirmation that it is on T0. 80 steps away from this place it has center of the disk. I really doubt that these mechanisms could be more sophisticated
I'm not sure. Maybe it's like a shock absorber, just to deal with the small vibrations. That's why earlier versions have a rubber stop inside the motor case instead of that tab. But yeah, not a very good long-term solution.
Hmm... while this isn't really useful to me, I have a related problem and perhaps you can give me some clues? I have an ancient Commodore 2031 floppy disk drive (5 & a quarter inch) which I was using for mass storage on an old kit-built computer because the GPIB interface (IEEE-488) is easy to connect to (I made my own home-brew interface for it). The trouble is that this drive seemed to always go out of alignment even though I can't find any "play" in the stepper motor. I'd format a disk and use it for a while and then, even not using the drive for a while, the disk would suddenly become unreadable. Reformatting would fix the problem for a while but the same thing would happen after a while. I'm interested in whether I can find a video on aligning and maintaining one of these old drives. (Edited to correct the disk size)
Usually when that happens is the same problem: the stepper motor isn't staying totally in place. Since there's no play, it must happen slowly over time. In earlier models there's a rubber ring inside the motor casing, and when it wears out, you get similar behavior to what you're seeing. Might be worth exploring that. Good luck!
@@NoelsRetroLab It sounds like that rubber ring might be the problem. That'll be a real monster of a job if it needs replacing. I might be better off simulating a FDD by using a PIC to interface to a SD card interface. But thanks for the info.
Other possibilities: 1) Maybe the disk’s heads need to be cleaned. Only use cleaning disks on double sided drives [Q-tips will destroy the top head] 2) Maybe the diskette itself is dirty or worn out. 3) Magnetic media degrades over time. Our floppy archive system required the archived disks be reformatted and re-written every ?? years [or was it replaced rather than reused?].
You should have marked the motor position with a marker before you removed the motor housing so that when you reassemble it, it will be very close to the original alignment.
8:36 3 in 1 is not a good choice here as it gets sticky and mostly evaporates due to having a lot of naphtha in it. white lithium grease or the grease that is used for 3d printers ( the name escapes me) is a MUCH better choice.
Yes, you definitely can. But I do like have physical drives whenever possible, so I try to save the Gotek devices as secondary drives (or even primary, as long as I still have a physical drive). So yeah, they're a hassle to repair, but it's worth it for me.
@@NoelsRetroLab Interesting! I like watching your channel cos I was one of the first people in the UK to own a ZX81, then later progressed to a Spectrum with external microdrives. Sadly, I don't own one anymore. The reason I asked the question is because I am now into vintage synths, one of which is a Roland W30 sampling workstation from the early 90s. It has a built-in 3.5" floppy drive, and the keyboard needs to be booted with a system disk inserted. These FDDs are notoriously fallible, and although mine seemed OK, the first thing I did was replace it with a Gotek, using Keir Fraser's (free!) FlashFloppy emulator firmware. All sounds/samples/system disks are stored on a USB drive, and with a rotary encoder, using the synth is now a breeze. I also use the SCSI port to talk to a device called a SCSI2SD (various versions available) which allow you to emulate 4 external SCSI hard drives on a single SD card.
I don't get it. At 21:00 when you adjust the motor a little, and try to read the disc again, I can see the read the read head going through its positions trying to read the disk. I have a NEC 386 laptop, and the drive head in mine doesn't move even a bit after adjustments. So pasically there is zero effect of me keep adjusting the stepper motor. Is there a Dos command to move the read head? I have no idea what I am missing?
What I'm doing off camera that maybe I didn't explain well, is I'm pressing N and P (for Next track and Previous track), just forcing the disk program to move again. Otherwise you're right, you wouldn't see a difference.
@@NoelsRetroLab I somehow managed to turn the drive into a position where it reads now. Pure luck, probably one in a thousand chance, I still can't believe it. I would have been completely screwed otherwise, as I had no way of putting any disk program on the computer without the floppy drive working in the first place. Which meant that after every single manual adjustment of the motor, I had to switch off then on the machine, so it would reposition the floppy drive's reading head. Is there any DOS utility that you know, and would recommend to test floppy drives?
![[TH] 2024 PMGC League | Survival Stage วันที่ 1 | PUBG MOBILE Global Championship](http://i.ytimg.com/vi/FVFT-PFwU80/mqdefault.jpg)
at 10:58, I have to say this is the BEST (and only?) illustration that I have ever seen on TH-cam of how a floppy drive head writes data to the physical disk. VERY well done! I'm creating a video that will cover an overlapping topic, and you have inspired me greatly here! Thank you!!!
This is not just a video, is a tech document. Your work is amazing, Noel !!!!! I trick I learned time ago to press that metal piece in position is inserting a small one twisted 90 degrees inside the same hole, between the piece itself and the motor structure (the round metal piece), in a way they make a "cross" and one pushes the other. Then, try to keep them together with glue or solder. It worked "some years" ago....
Thanks! That's a good tip! I'll keep it in mind for the next one I come across.
Brings back memories, back in 1984 I was repairing FDD at Mitsubishi on 8" floppy drives, then 5.25" and then 3.5". Radial and azimuth alignment was the norm. The key was getting the screws tight as shipping (trucks bouncing) could easily misalign the heads, specially on the 3.5", you need to start tightening the screws when the head was out of alignment, and tighten into alignment, most techs just could not get this right. Adjusting rotation speed, track 0, disk clamping force were all adjusted. I miss the Brikon FDD tester analyzer. www.techtravels.org/wp-content/uploads/2008/12/brikon723-1000x1.jpg Much more fun for me was repairing the circuit board to component level, thats why I moved to PC motherboards from the FDD department.
Good tutorial. I am currently restoring a completely busted MSX2 (Yamaha YIS-805). Both disk drives have broken/missing diskette presence and write protection sensors and have a lot of rust and oxides inside. I had to completely disassemble the entire disk drive and clean its parts from rust. After assembling back, the drive obviously needs re-adjustment. Unlike the drive you showed, the JU-253 drive has track 0 sensor that needs quite a lot of precision to adjust correctly. But it's doable even without the oscilloscope. What really helps is to mark the exact position of the parts before disassembling with either a marker or with a sharp object. This saves quite some time because after you align everything as before using the marks, only very minor adjustments are needed to make the drive work again.
Oh this one has a really fiddly track 0 sensor too! It's just that it was in the right place and didn't have to be adjusted. If both the track 0 and the motor alignment are off, it becomes quite challening. That will be a topic for another video I'm sure.
This is the exact video I didn't know I desperately needed. You've outdone yourself again, thanks a lot Mr. Llopis!
Happy to help!
Used these instructions to realign the heads on one of my drives. Brilliant work noel. thanks a lot.
So glad to hear that!
Great job! To help rotating the motor more precisely, and also have the screws more thightened while rotating (so the motor does not move when final thightening), you could 3D print a lever, and clamp it on the motor. This way, you have more torque and more precision. I will try to design something this week-end.
I mean, there already is kind of a "cogwheel" around the motor that's probably meant to be used exactly for that. All you need is a screwdriver.
Excellent video Noel! I have recently subscribed to your channel. I am currently restoring a CPC6128 as it was the first machine I owned as a child. I really appreciate you sharing your outstanding knowledge to help me and others fix up these old machines. Keep up the good work!
Welcome aboard! Glad you're enjoying the videos. Good luck with that 6128! Such a great computer.
I did the same thing, but on a 3.5" double-sided drive. I didn't use a computer in the process at all, instead I used another drive, to record a square wave pattern on both sides of dist at track 0 (using function generator), and a silence on track 1 (actually you can record even analog signals on the Shugart floppy drive). Then I used this diskette, to align track 0 and top head on the drive being repaired, using the oscilloscope.
Nice! I like that method a lot. I wonder if I can do the same thing on a 3" disk. That would make the oscilloscope method more reliable.
@@NoelsRetroLab I don't know the Amstrad floppy interface, but believe it doesn't differ from a Shugart/IBM. If so, then you can control the drive with a simple Arduino sketch, or even with some switches for direction and side select, and pushbuttons for a step signal and a write gate.
@@TzOk True. I looked at the interface and it seems pretty simple. That's a fun project idea!
Everything you upload is great. I love your content. Please keep up the good work!
Thanks so much, Corey!
I concur! I always look forward to Noel's videos!
I feel like a fixture that could snap around the motor, containing a spring that could sit on the back of it, pushing in on that little tab would be an easier fix, and could prevent it from happening again... cool vid!
I like that. It would have to be pretty thin to fit in the case, but it seems definitely more robust than bending the tab like I did or using some kind of tape like some people are suggesting.
@@NoelsRetroLab I think, the flimsy tab won't survive re-bending very often anyway. It seems like the motion of head assembly deforms the tab over time because it pushes it back. A bit more support for the tab should solve this problem. Look at the other drive model with the bigger backplate on the motor. How is it solved there? Great video!
While watching the vid I was thinking the same: a metal strip ~2cm wide, shaped like back of that motor, and held by the same two screws, and either a spring like Scott suggests or even part of that piece bent to reinforce the tab. I wonder if it can be improvised bending some old Meccano.
Interesting but why didn't you mark the position on the housing and the motor before you removed the motor?
... that was my first thought, too. At 10:37 you can see the washers still in place. But I think as an educational video its fine to explain the alignment preocedure in case you have to exchange the head- motor.
There really isn't any need to even remove the motor. Just slide a wire behind the tab and bend the font in as you hold it out. Then release and it should spring back forward again. Same issue happens to some 3.5" floppys for PC's. If you want to remove motor, mark the position.
You have no idea how much you helped me out with this video. I am a noob when it comes to computers, but I have an old 386SX laptop with a broken disc drive. It turned out that it was just slightly seized due to never been used. but I had no idea about this alignment issue with floppy drives, and needless to say my drive doesn't read anything now (I removed the stepper motor to grease things).
My issue is, it is a laptop, so I can't just simply remove the drive and fiddle with it. And disassembling and putting it back together like a hundred times after every adjustment is just not feasible. So what I am thinking is getting an extension cable, and route the drive outside of the laptop, and do those adjustments there.
The only cable though that I could find is an extension cable +board combo for Raspberry Pi. Also the cable has two female ends, so will need to do some soldering for a male connector. I will try my luck, as I don't see any other options.
thank you for posting. before disassembling the motor to bend that little metal tab, i might have tried just pressing on a postage-stamp-sized square of aluminum foil tape. if that little tab bent once, it'll probly bend again. al foil tape is available from home center or hardware store. good n useful instructions for stepper motor alignment. never tried that. i wonder what the little teeth around the base of the stepper motor are for. perhaps to resist rotation of the stepper motor housing ? please take care n stay safe.
I considered trying to tape it in somehow, but I was afraid that would come loose sooner than the bent tab itself, so I opted for this solution (and got to align the motor at the same time). But I'm sure that something could be done along those lines.
Another great video. I have learned how to align so I can adjust track 0 on these drives. Thanks.
I don’t think I’ve ever seen a 3 inch floppy drive before, so never worked on this type. I was a service tech for a big computer dealer back between 1979 and 1985 or so and did work with a lot of 5.25 inch and 3.5 inch drives though. We used to realign a lot of the Apple (Shugart) 5.25 inch drives and later some of the IBM full height 5.25 inch floppy drives (Shugart and Tandem). For the 5.25 inch drives there was a special alignment diskette available from Dysan that we used. It was actually recorded with the test tracks off center, so on a scope it showed what was known as a Cat’s Eye pattern. Basically the pattern looked like two different sized ellipses on the scope, and you adjusted the motor until the ellipses were equal in height. Your methods seem to work well enough though, you could probably even do that with a digital voltmeter set to AC instead of a scope, as long as it’s input impedance was high enough that it didn’t load down the signal. I realize why you went through the process as a teaching experience. For my own equipment though, I think I would’ve just glued something onto the back of the motor to hold in on the lever and avoided the realignment? Thanks for the video.
Cool. Someone else mentioned a special disk. Even a disk with all 1s in one track and 0s in another track would be very useful zeroing in to the correct alignment. I might try that next time.
You're right that there was probably something external that could do the job. I wasn't sure what would be permanent enogh (or at least as permanent and bending the dumb tab, which is not very), but it was fine since I also wanted to show the alignment procedure.
My memory is a bit fuzzy on this. I started out using the analog cats eye method and then switched over to a digital alignment disk. It was faster because you didn’t need to hook up the ‘scope. Don’t remember how the digital alignment disk was actually used. Couldn't find a source for the digital disk and only one New Old Stock analog disk on EBay.
I only used the analog diskette so not really sure how the digital worked. I haven’t looked through these yet, but got several hits just googling “Dylan digital drive alignment diskette”.
i thought that all cpcs had the same drive model but i was wrong ..... exellent video!
This helped me revive a 1992 Panasonic PC floppy disk drive. Thank you!
This brings back some memories of me fixing my amiga floppy after it stopped reading disc. Unfortunately i don't have a oscilloscope. But i got the discs to read. There were to brands one was sony and the other was chinon. Thanks for the video
Some great info here.
I never knew the motor just had a flimsy piece of metal to keep the spindle aligned.
Bit of a trick, btw...
You can actually hook up an audio amplifier to the signal from the read head, and "hear" the tracks.
It's not a very scientific method, of course, but if you don't have an oscilloscope, it can help you get aligned with the first track.
You can even doing this by hooking up to the /READ signal on the ribbon connector (instead of directly to the read head, where the signal voltage is tiny).
But, the /READ signal is +5V logic level, so if you hook up to an audio amp, definitely add a series resistor (or around 1K) and a small capacitor (say 22uF) for DC blocking.
Thanks! Only that cost-down version. Earlier versions have more of a rubber ring (which also ends up failing). But the whole drive feels very cheap.
@@NoelsRetroLab Yeah, that pretty much sums up the Amstrad build quality back then. lol
But I still loved my 6128 and Speccy +2 in the 80s/90s. ;)
I bought a 6128 a few years back which had a faulty drive. I replaced it with a Gotek, but it might be interesting to see if it has the same issue.
Although I no longer have any 3" disks to try, just the one that came with it.
This will help me with my Spectrum +3 Drive. Thanks 👍
No problem 👍 Glad it's helpful!
I have 2 suggestions :
First you can draw a mark on the motor before disassembling it, then you can put it back at the same position it was before..
Secondly, please use silicon based oil (or grease) on plastic.. Some plastic are "dissolved" by petroleum based products.
Agreed: use a sharpie to mark the orientation of the stator.
On IBM drives at least I always just made a reference mark on the stepper, then marked the low and high point where the drive would start to have issues and then locked it down in the middle of those two ranges. An oscilloscope of course is the best way.
Hi Noel greetings from Australia! Really enjoy your videos especially the troubleshooting ones. I'm an electrician by trade and really one day want to get into electronics myself, but to a dumb wire pulling electrician - electronics can seem like pure magic at times - especially vintage analogue circuits! One day your vids will push me into buying an oscilloscope 😂
Keep up the good work my man! I think you are doing a great job 👍
Thank you so much! It's not as complicated as it sounds, and you have most of the basics down! And honestly, vintage electronics are way easier than modern ones. Much easier to understand and diagnose. I didn't know real, practical electronics to repair computers 5 years ago (I came more from the academic side and the low-level programming side), so you can definitely pick it up and get started! Good luck with that!
hi, good job. Recently i repair a 3.5 floppy drive, this drive also haved alignment problems.
I've never done it with a 3.5 drive. Is it a similar process to realign a 3.5 disk drive, or is the mechanism very different? I imagine they still have a stepper motor and a worm shaft at their core, right?
Excellent video! I see a few people have mentioned fabricating some device to keep that motor tab pushed in, here is my suggestion. Find a non-functional audio cassette, remove and modify the metal bar and felt pad assembly that maintains pressure on the tape head. Cut down the felt pad to the right shape to sufficiently push in the tab . A slight lengthwise “V” may need to be bent in the metal bar to stop it from flexing. Solder the bar in place. I think the metal is copper or brass ... will that solder to the motor casing? A tiny bit of epoxy on the modified felt pad may be necessary to stop it fraying (dry completely before installation). That tab design is a really bad example of design cost reduction.
I suppose soldering some random metal plate would do as well…
Maybe a nickel strip with a spot welder?
Hello, this is a great video, thank you for posting it, it is very instructive. I left a mark with a marker pen to keep the position of the stator. and I was successfully able to CAT the dscs even after removing the stator, and that is still possible, as you mention, in several places. I have 2 questions 1) Does moving the ring that is screwed on the other side of bushing have any influence on the indexing? 2) My motor is only moving outwards, it cannot / will not move inwards, and I do not understand why as I "rewet" the solder positions of cable of the stator. I cleaned the shaft and the stator and it all rotates OK. I am inkling towards a problem between the stator and shaft but I am puzzled. Thanks in advance for any advice you may have. (FDD is an EME-155)
Well done on demystifying some of those fixes. So logical!
Thanks! That's the idea 😃
Nice video, I'm working on a floppy project at the moment, so this was a nice video to help me understand the stepper motor a bit more and now I know what the T1 and T2 points are!
Glad it helped, but tell us about your project! Is it specifically with 3" drives?
@@NoelsRetroLab Im trying to build a replica of the Amstrad SI-1 that was supposed to have been released for the ZX Spectrum +2A but never was, to add floppy support. I would love to use 3" but they are difficult things to come by, so I'm learning how to adapt an old 3.5" PC drive - really got into my Shugart spec and READY Vs DISK CHANGE signals! My wife thinks I've gone mad! Also though noone had made a SI-1 before, but then I found them on sale at 8bit 4ever (I think that's the name of the shop) and it looks really good, so maybe I'll buy one of those of it doesn't work out. Was soo jealous of your broken 3" drawer 🤣 so many drives... 🤤
That's awesome! Good luck with the project! As for the drawer... don't be so jealous. I'm not sure how many of those are fixable! 😬
@@mogwaay A lot of people just use 3.5" pc drives on the speccy +2A/3 as drive b, I used an old amiga external drive, I took the internals out followed a guide on making a cable with the ready signal, I also followed a guide on making the drive switchable from b to a so it can boot games and I cut it down to the size of the speccy +3 so it can fit on top of my +3 drive, I painted it and put a +3 sticker on, looks really nice, here is a link if you are in the SFEO group :-)
m.facebook.com/groups/spectrumforeveryone/permalink/927593487587123/
@@Digital-Vortex Hi Chris, thanks for that. I've joined the Facebook group, not heard of it before, so thanks for the tip. I finished wiring up my prototype SI-1 clone and it seems to work - my +2A thought it was a +3 and said drive A: was available - woo hoo! At the moment I'm just separating and grounding the ready line at the moment from the drive and ignoring Disk Change from the IBM 3.5" drive and the Speccy seems ok with that. Unfortunately, at first attempt (at part-my-bedtime o'clock!) the floppy drive wasn't reading the disk (couldn't find track 0) or letting me format, but it did seem to be trying, so I'm taking that as a win for now and will look into it when I get next chance. Cheers!
My new favourite TH-cam channel.
I've been reading your Amstrad diagnostics tools on GitHub. What a contribution!!
I'd love to know however where can I purchase the dandanator?
I am surprised they don't have a self alignment feature. I am used to steppers on CNC and laser cutters , and the "home" when they start up . great bit of info , thanks
Good point. I guess in order to self-align the motor needs some way to rotate in small increments. I suspect that's not possible to do from software (otherwise you could run some sofware that would mess up your drive), but it would be very interesting. I don't know if it's a limitation of the controller or the stepper motor itself. If it was just the disk controller, I could imagine trying to write a self-alignment program with an Arduino. Hmmm...
@@NoelsRetroLab with CNC ETC , you have a hall effect or other sensor to detect the "home" position
@@outbackchillin5628 Here too, there is a home sensor: the optical one, on track 0. But the resolution of the stepper is not enough to move between tracks, perfect alignement is needed between the steps of the motor and the tracks position.
@@NoelsRetroLab I think these old small motors only have a few full steps per turn (modern hybrid steppers has 200 or 400). Modern steppers also allow drivers to do micro-stepping, by not energizing coils at zero/full current, but in between, so the stepper does partial full step rotation. On the other hand, old stepper motors don't allow micro-stepping, and only turn half steps. This is old technology, so nothing much to do.
Another great video Noel! I don't have a 3" disk drive or even a CPC, but I enjoyed it anyway. I agree, that design is ridiculous. Someone was pinching too many pennies.
Also, happy channeliversary! I didn't realize you've only been doing videos for one year. They are very well organized and made. I look forward to the next one and many more!
Thanks! Yeah, typical Sugar I'm afraid. Always looking at the bottom line. I'm actually amazed that the Amstrad CPC turned out the way it did when that was the driving mentality at the company. And thanks for the anniversary wishes! Crazy! It seems like a long time and not very long at the same time 😃
It's odd that the outside diameter of the motor has gear teeth, maybe there was a tool like a screw driver with a small gear on the end and a pin to fit in a hole on the motor mount.
I have done the same tweaking you did to fix the motor position but 30 years ago, but on the ancient Tandon Tm100 5.25 inch drives. CLUNK CLUNK BZZZZZZ lol.
Very old but fun to fix, and they had a strobe disk on the disk hub wheel so the speed can be set.
The plus D disk drive interface for the speccy worked on old and new drives.
What a great explanation of how to check/repair the 3" drive. I imagine that it would be the same for a ZX Spectrum +3.
So, my question is, is there also some test to use on the Dandanator for ZX-Spectrum, like the one u use for the Amstrad. Thx :)
Test for moving the head for the disk? I'm not aware of any, but maybe. In general the Spectrum didn't get nearly as much disk support as the Amstrad did.
I don't have any special knowledge or anything, but one thing did occur to me about about aligning without an oscilloscope - you won't know if you've actually aligned to the centre of the track. So perhaps it could work on most disks, but then you might write to a disk and have it not read correctly on another drive, or find some disks don't work and assume they're bad, because those drives or disks had their own slight misalignment which might have been within tolerance for a correctly aligned drive but not for your slightly misaligned one.
Not sure. I'd have thought using a scope would be best if you have one.
I would have tried a relay adjustment tool to bend the tab in before removing the motor. The tool is a slim rod or bar with a slot on the end. Put the slot over the tab at the attached end and twist to force the loose end against the motor shaft. If it didn't solve the problem you still have the option to remove the motor.
Interesting video. I have the same issue with a 128 setup I got given last year to add to my collection. I'll try the metal strip and if that fails, the motor as I've already changed the belt. It's like aligning old radios which I don't tend to do as I don't have an oscilloscope but the slower method, I'll have a go at. Thanks ps - subbed.
Thanks, glad it was useful. One thing to watch out for: Do you get Drive A: before the error when you type CAT? If you don't, then it's likely the track 0 sensor. If you do, then it could be an alignment problem (or a head problem, or a data communication problem). So always start there.
@@NoelsRetroLab thanks for the reply. I'll dig it out this weekend and let you know how far I get
When I did something like this decades ago, I thought it was important to align it with track zero starting from the edge where there is no signal at all, otherwise you could accidentally align it to an adjacent track making every track one off the correct track. Perhaps the software is verifying that it's reading the correct track?
Interesting video, thanks for uploading.
oh
and
by
the
way
...
Happy 1st anniversary on your channel, Noel! 🎈🎁🍰
Thanks! 😃 Wow, someone noticed! 🎉🎉🎉
@@NoelsRetroLab By the way, it occurred to me after i had posted my comment that it might be helpful to mark the position of the stepper motor with a fine-tipped sharpie before removing it - i.e. a mark that covers both the motor and the chassis; that should make it easier to realign it afterwards, don't you think?
@@BertGrink Yes, I think it could be helpful as a starting point, but it's so fine that even that will need to be adjusted. But it seems like a good thing to do anyway. Next time I'll try that.
@@NoelsRetroLab Please report back how it works out, if it's not too much of a bother.
I wish I knew this kind of stuff when I was younger. Thank you so much for sharing and demonstrating!
Glad you enjoyed it!
I have to agree - it only took me an additional 30 years to "figure" some of this stuff out. Mainly because of lack of access to certain information.
Yeah totally agree, would've loved to know this stuff when I was a teen, but alas it was just so hard to get information like this... It's great to come back to it now, have been really inspired by Noels videos to give it a go.
Noel, if I was to detail the full procedure to you and how to set up TS00 after adjusting the Stepper, would you be interested in making a new video? I’m sure you could use this video and cut in new footage. Also checking the index pulses at 200ms. Personally I think this would give the full story for 3” drives and how to set them up post service with lub and belt.
Hi thanks for that nice explanation ,I did it and it worked great but not on all my disks is it because there used ,or do I have to try again?
Nice. Since you had a partially working drive, would it have been possible to measure the voltage before removing the motor and then you would have known roughly what you were aiming for when you reattached?
Good idea. I should have done that! (or measure it now). I'll probably do another video on these drives at some point about adjusting the track 0 detection, so maybe I can do a quick followup addressing things like this. Thanks.
Great job! I would have really just took a chunk of plastic and fixed it so that tab was solid against the back by gluing the plastic to the back or something like that. lol
Hi, I have a 6128 FDD that sometimes fail to read discs, and sometimes also randomly erase CAT tracks when I just want to erase one file. I'm going to apply your adviced to try and fix it (from this video and the first one too).
But my drive also have an issue that I don't think you've covered in your videos. When I push the eject button, the drive sometimes fail to push it out, and I have a hard time to remove it. I guess maybe a spring is to weak to do the job ?
This could be a savior for me,I have 2 drives with same error,I'll try to fix it!
Awesome! Let me know how it goes. Or drop by the Discord channel if you need some help.
Thanks for another great video, I like the way you do the graphics explaining what's going on with the sectors/tracks :-) quick question though, Is it possible to use a multimeter instead of doing it by trial and error or oscilloscope? Also is there any software to help aligning for the spectrum +3?
Thanks :-)
Thanks! I'm not sure you could use a multimeter, although Mr. Guru in a comment above is explaining that you can, at least on Amigas, so check out that comment if he writes back. I'm not aware of any software for that on the Spectrum, but there has to be some. If not, you can always move the drive to an Amstrad, do that, and move it back 😃
How about using an Amstrad emulator and a Greaseweazle? :-)
Thanks for great educational content... on legacy system hardware
Thanks for watching!
Hi! Thank you for this video it really helped understand a lot of things. I have a question regarding the test room. I purchased a M4 board from Duke and wanted to know if I can load a test room there and where to download the rom from?? Thank you again!
Hi Noel hope you are keeping well. Another great YT vid to enjoy :-)
Glad you enjoyed it! 😃👍
Good job!!I I just received an amstrad cpc 6128 today, and when power on the floppy disk is always with the led turned on. Do you think my problem is the same?
Thanks! The drive LED is always on. It just happens to get brighter when you're reading from it. So that might be fine.
Good job, video is amazing. Love 80s
Glad you enjoyed it
Ah that was an interesting video Noel ! and it's given me some ideas too, when i eventually get all my source code for Disced finished I'll produce a version that can be used for just setting track/sector(s), to help with testing this type of adjustment, that would be cool ! thanks for yet another great video :-)
Thanks! Check out the software in the description. It pretty much does that already.
@@NoelsRetroLab ah i missed that, but after having tried that util, the UI is awful, so i will definitely produce my own version of it.
Actually, someone in another comment had a great point: The oscilloscope method would be a LOT better if the drive were constantly seeking to a particular track (instead of having to tell it move forward, move backwards). That way you can see your changes in almost real time. So there's your unique angle for that software. I would totally use it!
@@NoelsRetroLab gotcha, i'll return to this someday when I get the code all documented and working !
Excellent video
Glad you liked it!
If the drives works well when pushing the tab on the back of the step motor, what about marking precisely the position of the step motor before removing it.
There are many "teeth" on the step motor, so just by putting a dot on one of them and tracing a thin line on the chassi corresponding with the choosen teeth tip, it would be possible to put it back just in the right position?
Of course this could only work if the drive reads correctly when the tab is pushed, or else it means the motor has been removed before and not align correctly when put back in place.
But if this is the case or you just forgot to mark the step motor alignment on the chassi, this method is very useful and works great! :-]
amstrad did everything on the cheap and this drive shows that Noel (piss poor design) nice tutorial Noel
Yeah, definitely. Quite a shame because their CPC range was awesome.
You can complain to Matsushita as they are the designers of the drives, not Amstrad.
Usefull information. Thank you. Luckily, I don't have any drives of that model.
Thanks! If we're talking about 3" drives, the other models have similar issues, different solutions. The motor is held in place with a kind of rubber ring which tends to become much after many years of use. No tab to bend in there to fix it, so you need to get creative. Fortunately those drives are sturdier and that doesn't happen much.
i wonder if they make really small c clamps with rubber pad on the screwy end to just clamp around the motor or 3d print a clamp from polyethylene to push on thecasting flat and go around and push the shaft spring
That's exactly what the earlier models have inside the motor casing! Eventually they degrade and become a mush and the drive goes out of alignment too.
Realizing that putting it back into alignment was the focus, since all you did was push the tab in, was there a reason you couldn't just push the tab in while it was still attached, if you didn't want to go through the process of realignment? Or is there also a technical/electrical reason you didn't want to do that?
Good question. I didn't explain that very well. Here's the situation: That's a springy metal tab and it's 1mm away from the motor. You can push it all you want, but you won't be able to bend it in so that it's rest position is past the motor (so it pushes slightly against it). See what I mean? You really need to push it past that point and then have it back into place.
@@NoelsRetroLab Ah ha. No, that makes sense, I just had a mental model of it that was slightly off of reality. I wasn't taking into account the inner cylinder (at 10:20).
When you said it was the magnet that was rotated I didn't think it had a direct impact on the tab. Looking at it again, with your comment in mind, now I have a better idea of what the inside looks like.
Thanks for the reply! :)
I found an old drive that the head assembly is loose from the rail, so the motor spins but the head barely moves. You can move it back and forth with your fingers without the rail moving at all. I'm not sure if there's any way to fix that?
Great video Noel
No words! Thanks!
The advantage of the oscilloscope would be to monitor the waveform in realtime as you rotate the motor. Without having to restart the read every time. I'm not familiar with the CPC, but I expect there's a way to command the drive to read continuously, or even if that's not possible, just write a simple loop which reads a certain track over and over.
Very good point!! It would be easy to write such a program too! Thanks for the tip.
@@NoelsRetroLab and use an analog (needle) voltmeter, a sensitive one like the ones used for aligning vintage radios..watching analog needle movement is perfect when searching for peeks..much better than scope or dmm
@@NoelsRetroLab But how the signal can be refreshed in real time (as we can see jiggling on the oscilloscope), but not been really refreshed regarding to the amplitude? Is there some sort of buffer, holding it? I'm puzzled...
Hi, could you do a video of adjusting the azimuth on the Amstrad CPC 464 tape deck please? I managed to record an azimuth adjustment program to tape and it helped slightly but games still fail to load, I'm looking for an accurate way to adjust the azimuth. Thank you.
Added it to the list! Thanks for the suggestion.
thats interesting.. i remember changing the head in my 464 back in the day and never needed to align it in any special way.. but maybe i just got lucky!
i also 'made' one of them tapes with the head in it and a stereo cable comming out.. and never ran into alignment problems with it playing from a stereo system.. (dont ask me why i made it or was using it that way...i was just screwing around!)
@@WacKEDmaN Haha yes, you definitely got lucky!!
@@NoelsRetroLab Thank you.
Thanks for the interesting video! I noticed that there was a ball bearing hanging from the worm gear at 11:52 did you put it back in? :D
good video, I was thinking, if that failure point on the stepper motor has got lose with time, it might not be better to mod it? you could easily weld a metal strip with a spring, to keep it always tight
"weld a metal strip" and blow a huge hole in the cheap as tin can!
@@WacKEDmaN maybe I used the wrong word, solder, thin should be strong enough to hold something in place and it can't blow a hole anywhere
@@danielhaupt2066 yeah true.. i forgot some ppl use weld and solder interchangably! :P ..needs a plate like the other models.. im guessing they have a spring to push against the shaft..
Yes, that's probably a more permanent solution. Next time I might try something like that.
Hi, do you have documentation about those drives? I'm trying to properly emulate them for ACE emulator. Thanks
Thanks for the video =)
My pleasure!
excellent content Noel!.. as usual!..
..that stepper looks soo damn cheap! (more sweet Sugar goodness!) i was amazed the windings come away from the shaft like that..but i guess its not that bad of a design considering they had to align the head somehow...
Yeah, this drive feels flimsy and cheap compared to the early drives! Even the stepper motor is 30% smaller. But hey, that's what Sugar was all about.
THANK YOU ...... I DID IT ..... THANK YOU AGAIN
What if you use a felt pen to mark the motor's alignment before disassembly? You could reassemble it more-or-less the same way it was after.
Hi, at 4:26 you speak abouth drive models on the CPC forum. I've searched like hell and didn't find that posts. Is it possible to place a link to that specifique page please. Thanks :)
Found it, was not on forum but on a WIKI page: www.cpcwiki.eu/index.php/Amstrad_FDD_part#EME-156
Right! Sorry, I only saw your question now. That's the page I was referring to.
Is it not possible to mark the correctly aligned motor position using a marker pen next to the securing screw?
Should've called this episode "What are you doing, step-motor?!".
what about 3d printing a cup to cover the back of the stepper motor and let the cup press le metal leaf? the cup could have a screw to press fit the cup to the motor and maybe a small dent at the center to better press the metal leaf
Thanks!
Great video!
Thanks!
I once realigned a 720K 3" drive from a PCW9512, using the trial-and-error method. It's time consuming, particularly because the adjustment of the motor is so sensitive - you have to apply pressure to it, but if you feel it move, you've gone too far.
Yes! 1 degree is probably too much to move at once. You saw I spent quite a while doing it here!
why couldnt you have just put some tape or something on the back of the motor to hold the tab in? must have been a reason that i dont understand
I considered that, but how would would a tape have lasted putting the right pressure? I thought it would be better to do it right. But maybe that would have been enough.
@@NoelsRetroLab or solder something on the motor case to prevent the tab from coming back out again
@@fft2020 Yes, very possible. Add some rubber bit to push the tab in and the solder a metal piece to the casing holding everything in place. That would probably work.
@@NoelsRetroLab yes, it looked to me like the rear of the motor on the later version that you briefly showed us did in fact have some kind of plate affixed to it’s rear instead of that tab. As you say; dreadful design for that bit.
Nice Noel.
Thanks you!
Do you have a URL for the list at 2:58....I have a Matsushita EME-213YV Im trying to repair
...ok, you got me! Imma subscribe now. Keep it up.
Welcome aboard!
Can you show us how to align a pc 3.5" disk drive. I can't find any documents or software.
please?
Thank you for the great vid, would it help to mark the motor and the chassis of the drive with a marker pen 1st before taking it off 🤔 and I have a few drives here but 1 has a white plastic front and 2 heads, 1 on top 🤔 does anyone know what machine it’s off. Thanks.
Yes, that'd be a good starting point. You probably have a 3" drive from an Amstrad PCW: 2 sides and white front. Very similar otherwise I think.
Thank you
Thanks for the video. I see... Great work.
That tab is maintaining pressure on the motor drive shaft. It looks like it's not quite up to the job. I would add something to the back of the motor to assist the tab. Otherwise the tab will probably go slack again over time?
But, this video is far more informative than just a bodge fix.
Yeah, I suspect it'll eventually go out of alignment again. Not sure how long it'll take though. Next time I'll try with a rubber ring on the inside, like earlier models.
7:15
i am wondering why to realign anything in the first place.
If you have track 0 detection this means to me that drive at boot doesn't know where it's at, so it backs step at the time until it gets confirmation that it is on T0. 80 steps away from this place it has center of the disk.
I really doubt that these mechanisms could be more sophisticated
Eres mi idolo camarada!!!
🤣🤣
It looks like this "tab" was designed (maybe?) to protect head from jams/locks, but it's clear that it works only one time :)
I'm not sure. Maybe it's like a shock absorber, just to deal with the small vibrations. That's why earlier versions have a rubber stop inside the motor case instead of that tab. But yeah, not a very good long-term solution.
Does this work with 3.5 mm floppy drives?
w\With the motor a ghetto fix would be to use some blue tack and very strong tape on the rear tab to push it in
Hmm... while this isn't really useful to me, I have a related problem and perhaps you can give me some clues? I have an ancient Commodore 2031 floppy disk drive (5 & a quarter inch) which I was using for mass storage on an old kit-built computer because the GPIB interface (IEEE-488) is easy to connect to (I made my own home-brew interface for it).
The trouble is that this drive seemed to always go out of alignment even though I can't find any "play" in the stepper motor. I'd format a disk and use it for a while and then, even not using the drive for a while, the disk would suddenly become unreadable. Reformatting would fix the problem for a while but the same thing would happen after a while.
I'm interested in whether I can find a video on aligning and maintaining one of these old drives.
(Edited to correct the disk size)
Usually when that happens is the same problem: the stepper motor isn't staying totally in place. Since there's no play, it must happen slowly over time. In earlier models there's a rubber ring inside the motor casing, and when it wears out, you get similar behavior to what you're seeing. Might be worth exploring that. Good luck!
@@NoelsRetroLab It sounds like that rubber ring might be the problem. That'll be a real monster of a job if it needs replacing. I might be better off simulating a FDD by using a PIC to interface to a SD card interface. But thanks for the info.
Other possibilities: 1) Maybe the disk’s heads need to be cleaned. Only use cleaning disks on double sided drives [Q-tips will destroy the top head] 2) Maybe the diskette itself is dirty or worn out. 3) Magnetic media degrades over time. Our floppy archive system required the archived disks be reformatted and re-written every ?? years [or was it replaced rather than reused?].
Please help me how to format 720 kb floppy disc
The motor case has gear teeth on it. Aren't they intended for alignment?
Not as far as I can see. Or maybe for the assembly or some machinery they used for alignment, but they don't correspond to anything on the other side.
You should have marked the motor position with a marker before you removed the motor housing so that when you reassemble it, it will be very close to the original alignment.
Yes, that would have definitely helped find a good starting position.
8:36 3 in 1 is not a good choice here as it gets sticky and mostly evaporates due to having a lot of naphtha in it. white lithium grease or the grease that is used for 3d printers ( the name escapes me) is a MUCH better choice.
Is it not possible to replace these 3" FDDs with Gotek USB emulators? Surely that would eliminate all this hassle?
Yes, you definitely can. But I do like have physical drives whenever possible, so I try to save the Gotek devices as secondary drives (or even primary, as long as I still have a physical drive). So yeah, they're a hassle to repair, but it's worth it for me.
@@NoelsRetroLab Interesting! I like watching your channel cos I was one of the first people in the UK to own a ZX81, then later progressed to a Spectrum with external microdrives. Sadly, I don't own one anymore. The reason I asked the question is because I am now into vintage synths, one of which is a Roland W30 sampling workstation from the early 90s. It has a built-in 3.5" floppy drive, and the keyboard needs to be booted with a system disk inserted. These FDDs are notoriously fallible, and although mine seemed OK, the first thing I did was replace it with a Gotek, using Keir Fraser's (free!) FlashFloppy emulator firmware. All sounds/samples/system disks are stored on a USB drive, and with a rotary encoder, using the synth is now a breeze. I also use the SCSI port to talk to a device called a SCSI2SD (various versions available) which allow you to emulate 4 external SCSI hard drives on a single SD card.
I don't get it. At 21:00 when you adjust the motor a little, and try to read the disc again, I can see the read the read head going through its positions trying to read the disk. I have a NEC 386 laptop, and the drive head in mine doesn't move even a bit after adjustments. So pasically there is zero effect of me keep adjusting the stepper motor. Is there a Dos command to move the read head? I have no idea what I am missing?
What I'm doing off camera that maybe I didn't explain well, is I'm pressing N and P (for Next track and Previous track), just forcing the disk program to move again. Otherwise you're right, you wouldn't see a difference.
@@NoelsRetroLab I somehow managed to turn the drive into a position where it reads now. Pure luck, probably one in a thousand chance, I still can't believe it. I would have been completely screwed otherwise, as I had no way of putting any disk program on the computer without the floppy drive working in the first place. Which meant that after every single manual adjustment of the motor, I had to switch off then on the machine, so it would reposition the floppy drive's reading head. Is there any DOS utility that you know, and would recommend to test floppy drives?
Maybe add a index mark before removing the stepper motor.
Good idea. It would help to have a starting position.
As you said: "what a horrible design"! Incredibly cheap indeed!
Seriously! The earlier models are more sturdy, but this particular one... sheesh!