News:

Printed Amstrad Addict magazine announced, check it out here!

Main Menu

Why, oh why, does this 3" drive run at 360 rpm?

Started by Spirantho, 16:50, 17 July 12

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

Spirantho

Afternoon chaps,
I was the lucky winner of an Amstrad CPC 6128 off eBay the other day. It was sold as not working, as the disk drive was reporting no disk. Fair enough, thinks I - it just needs a new belt.
So I take it apart and there's still a belt there. OK, that's a bit odd - must be slipping. Remove the belt and see loads of black gunk like drive belts do when they munge themselves all over the insides of your disk drive.
One load of swabbing and cleaning, and it's mostly clean. Re-attach the belt. Fire it up - "No disk".
Hmm.
Then I notice the 300 Ohm resistor for the index hole has become a 66.6KOhm resistor. One quick change later...
"Read Error".
OK, getting there. Re-align the head.... nope, no difference. Try everywhere... no difference. OK, it's something else.
Now some of you may be aware that I write the Catweasel drivers for AmigaOS 4. This is a Good Thing(tm), because it means I have all the low-level code for the drive interface, so I plugged in my drive, and got my drivers to dump out the raw pulse lengths.
Odd - they seem fine. Let's compare....
In goes a known good drive....
Strange, the pulse widths are different. But they're both reading the same pulses, just at different rates.
That's when I go back to my code and I realise the pulse widths look familiar - so I force it to use the timings for a 1.2MB floppy drive - which you may be aware are the only drives that run at 360 rpm instead of 300 rpm. This makes the pulses shorter when you use a 360K disk (from a 300 rpm drive) in a 1.2MB drive (at 360 rpm) as it's running 20% too fast.
And oh look - the disk reads perfectly.
So how on earth did my standard EME-156 drive in this CPC 6128 start spinning disks at 360 rpm? Is there a jumper somewhere? Obviously my poor 6128 can't keep up with it at this speed.
Can anyone help? Will be dead impressed if you can. :)
Thanks to anyone who can give me a pointer!
Ian Gledhill
Mutant Caterpillar Games Ltd Retro Store for repairs and sales of 8-bit and 16-bit retro systems and software

TFM

If you have FutureOS in a ROM box, then you can use it to format floppies in your drive. It's routines should be fast enought to work with it. Now, the reason for doing that it the following: If you can format a disc and read it (for example read directory - even an empty one) will show you that the drive works well on the CPC. In this case your discs are not readable for the presented reasons. However, it would be interesting to do that experiment.

Now, you changed that resistor right? Does this cause the speed up?

Or is it a 3" drive from Joyce PCW ? (B drive there is 80 Tracks, DS).

TFM of FutureSoft
Also visit the CPC and Plus users favorite OS: FutureOS - The Revolution on CPC6128 and 6128Plus

arnoldemu

I'm not sure the speed of the code would be the problem, would the problem not be more of the PLL?

If that can't sync to the data, then the fdc is not going to see the correct values, and it will miss the syncs and report errors or it will not see sectors... just as you have seen....

as for reading 300rpm discs, it can be done at 32us per byte, modified reading functions may do it (such as in futureos), but not standard amsdos ones.
not sure if it could read 360rpm though.

could it be that the drive was repaired badly because somebody didn't understand about replacing the belt?

the other thing is that 3" drives are not that great at rotating at 300rpm anyway. 3.5" are great because of the magnet and the direct drive, but because of the belt, the 3" drive doesn't run at exactly 300rpm.

really the answer is "don't know".
My games. My Games
My website with coding examples: Unofficial Amstrad WWW Resource

CanonMan

I'd be interested to hear if it is possible for 3" drives to suddenly start spinning at the wrong speed.

I've got a few drives which have been thoroughly cleaned and serviced and still won't read disks. How easy is it to get hold of a suitable test disk, to align the index sensor and head?

TFM

Well, as said: My idea would be to format a disc with that drive, then check if the directory can be read. If yes, the head it not on track or the speed is wrong, BUT the drive itself is working.  :)
TFM of FutureSoft
Also visit the CPC and Plus users favorite OS: FutureOS - The Revolution on CPC6128 and 6128Plus

Spirantho

Thanks for all the replies! This forum really is one of the best I've found on any platform for helpful and knowledgable people!


I know the drive is working because I could dump a disk successfully when I told my Catweasel to use 360rpm timings, but I'll try it on a CPC if I can anyway. I don't think it'll work though even if I can get a FORMAT program on it (I have a Multiface, will that format?) for the simple reason that if the disk is running 20% too fast then that means there's 20% less space on a track that then CPC is expecting as the CPC will continue to use 300rpm timings.
I've measured on three different disk drives that on one (an older EME-156 with the square chip) the motor gets 10.7V, and on this one and the other EME-156 which donated the resistor (which are the newer type with a fine-pitch DIL) the motor gets 12.1V. A difference in voltage on the DC motor (it's the same type of motor on both the later 156's) may change the speed, but then why is the donor EME-156 working?
I think the next plan is to put the resistor back onto the donor and see if the voltage goes down to 10.7V - not sure why it would, but I don't have schematics for this model EME-156, only the older type, so I don't know the circuit properly.
Good job I like a challenge, innit? :)
Ian Gledhill
Mutant Caterpillar Games Ltd Retro Store for repairs and sales of 8-bit and 16-bit retro systems and software

arnoldemu

Another thing I have heard reported is that there are 2 different sides to the belt itself. It could be fitted with the wrong side and cause slipping.
I would expect that to manifest itself as errors and not maintain a steady rate.
My games. My Games
My website with coding examples: Unofficial Amstrad WWW Resource

Bryce

On the EME-156 the 12V to the motor are switched by Q1 (2SD1198, a general purpose  NPN transistor), which is fed by an invertor directly from the Motor-ON signal. It also has a 1K resistor between it's base and collector. If the 1K resistor (R32) is missing / damaged / disconnected / wrong value, this could cause the motor to turn at the wrong speed. Check that R32 is ok. If the resistor is ok, I'd suggest replacing the transistor, although it is unlikely that the transistor has lowered its internal resistance without dying completely.

The M51017AP Controller IC has no connection to this motor, so the fault can't lie there. In fact no other parts are responsible for the motors speed, so it can only be one of these two components (or the motor itself has been replaced with the wrong type).

Bryce.

Spirantho

I sorted it!
It was....
.. the motor!
I don't know why, but this motor (with the same part number on it as my other drive) rotates at 360rpm instead of 300.
Thanks for all your help - including the extra bits that Bryce said - I may not have needed that knowledge after all but it'll likely be helpful in the future!
I've posted about my adventure on Mutant Caterpillar Games – The Blog!.


Many thanks again to all who helped!
Ian Gledhill
Mutant Caterpillar Games Ltd Retro Store for repairs and sales of 8-bit and 16-bit retro systems and software

Bryce

If it's exactly the same motor, with the same part number, but spins at a faster speed, then the coil inside has a short across the windings, so bin it.

Bryce.

arnoldemu

@spirantho:in your blog you say you need a motor?
I've got a couple I've extracted from drives and never used.
can't say how good they are.
want one to try?
My games. My Games
My website with coding examples: Unofficial Amstrad WWW Resource

ralferoo

I think I can explain why the resistance was too high... a good chunk of the resistor appears to be missing!

Bryce

I don't think there's anything missing from the resistor, it's just looks like it's soldered in upside-down. And even then, a chunk missing from a resistor would mean it could only carry a smaller amount of current, it shouldn't effect the value of resistance. However, the fact that it's upside-down would suggest it had been replaced in the past.

R19 should be about 180R as far as I know. It's the current limiter for the Track00 LED (not the Index LED as you suggest in the Blog), but you couldn't measure it in position, because other parts in the circuit could cause a false reading. For example, putting the meter across R19 would be measuring R19, but also be charging / discharging C22 at the same time (using the power from the multimeter), which would result in the value being higher or lower than it actually is.

Bryce.

P.s. did you measure the value of the resistor after you removed it?

endangermice

#13
I've been following this post with interest, Ian another great write up - keep them coming, I love reading these!


Looks to me like that resistor must have been changed at some stage and it does indeed look like it was soldered in upside down. I'm trying to think why it might have been replaced and my guess is that the original got knocked off the board somehow, perhaps while the previous owner was botching a repair. I suspect in the process they lost the original and not knowing much about electronics found another similar looking component on another board and just soldered it on (soldering skills seem surprisingly good). They then decided to replace the motor with one from another drive they had lying around not knowing that it ran at a faster speed and voila - one broken drive! Without specialist tools fixing the drive was going to be a matter of pot luck which didn't bear out on this occasion!


I'm collecting a CPC this evening I bought on Ebay at the beginning of the week, though sadly this one has been listed as working - still I can hope that the seller was lying and it will need fixing ;).


Ian, I may have a spare drive motor coming my way at some stage soon. I'm negotiating to swap one of my spare Archimedes A3000s with a French 6128 Plus. It has a broken 3" drive so I'm intending to fit a 3.5" drive rendering the existing one redundant (I don't have any 3" discs so keeping that drive is pointless). If the motor looks good and is suitable I'd be delighted to give it to you. We'll be hopefully finialising the arrangement soon and then have to wait for delivery from France etc. so may not be for a few weeks but I'll let you know if I get a working motor!
For all the latest Starquake remake news check out my website - www.endangermice.co.uk

Bryce

The 3in drive in the Plus uses a 5V motor, not a 12V like the EME-156 does. But I would try fixing the Plus drive anyway instead of using it for parts, they are much harder to find than standard CPC drives.

Bryce.

Spirantho

It's not upside-down, no - the back is white like a normal SMT resistor. It's just melted a bit I think. Whatever it is, or was, it's now a 66K7 resistor. Useful, huh?
@Bryce
R19 is definitely for the index hole, though - I can trace it on the disk drive itself - R19 connects to the anode of D2, the index LED.  Maybe they changed the numbering for the different revisions of PCB - this one is part number Z80264. The charged capacitor wouldn't increase the resistance, anyway, it'd lower it, because you can't increase a resistor's resistance by putting something in parallel with it.
@ArnoldEmu
A motor would be awesome, how much do you want for one?
@Endangermice
Thanks for the offer but NOOO! Don't do that! :) The 6128 Plus uses a 5V drive instead of 12V and they are rare as a hen's back molar. If you don't want the drive though, feel free to send it to me and I can try and resurrect it! Alternatively, maybe I can fix it for you? I know 3.5" drives are much more practical but somehow they don't have the same je ne sais quoi, you know what I mean? I like things to be original. :)
Ian Gledhill
Mutant Caterpillar Games Ltd Retro Store for repairs and sales of 8-bit and 16-bit retro systems and software

endangermice

#16
Ah I see, thanks for the heads up guys, having never seen a Plus in the flesh before, I know very little about them. I may well hold onto the Plus drive for now then. I'm guessing it's similar to the ones fitted in the later PCW (9512) computers apart from the fact that it only has one head?


I like to keep things original too when I can but due to the fact that French Pluses seem to be rather more common than their UK brothers I figured modifying it for a 3.5" drive wouldn't really matter - though if you guys think I'd be better off keeping it original (i.e. holding its value) then I might well be up for getting you to fix it Ian. Would love to do it myself but i just don't have the tools required for head alignment. I will give it a quick once over though - it might just be a broken belt.


I'm really buying the Plus because I want to play around with coding for it - WinApe is a great emulator but there's nothing like having the real thing, with the perfect 50hz sync to a monitor so I figured fitting a 3.5" drive would make it much easier to use for its intended purpose!


At the end of the day it really depends on the condition of the machine it's described as good condition apart from the dreaded ABS plastic yellowing but if it's a bit sub par, the modification is less of a worry!
For all the latest Starquake remake news check out my website - www.endangermice.co.uk

Bryce

The Capacitor doesn't really increase the resistance, it just pulls current from or feeds current to the meter, which makes the meter show a false reading (which can be higher or lower than the actual value). In reality, yes you are correct, nothing in parallel will increase the actual resistance.

The diagram I use is this one. R19 to Track00, R28 to Index (and it's 300R as you suggested), so obviously the numbers were changed at some stage.

Bryce.




Spirantho

Yes, that's the one I have from the 6128 service manual. That's for the one with the square ASIC or whatever it is on, this one is the more recent version.
@endangermice
The 5V drives as far as I know only exist in the 6128+. That's why you should treat yours with due care and attention. :)
Ian Gledhill
Mutant Caterpillar Games Ltd Retro Store for repairs and sales of 8-bit and 16-bit retro systems and software

endangermice

I shall take your advice and make sure the drive is treated like a princess. I actually got 3 x 3" discs from my latest purchase today, an immaculate 6128 - so now I have a reason to get those drives up and running!


This one seem to have the usual drive belt issue - however is there's still no dice after fitting a new one, this thread and Ian's blog will give me a whole host of things to try!
For all the latest Starquake remake news check out my website - www.endangermice.co.uk

steve

If you can get the 3" drive working you can hook up an external 3.5" drive, allowing you to make 3" disks to test new software on a real CPC if you have one.

endangermice

That's a good point! I've built an external 3.5" drive enclosure for my 6128 - so just need a convertor for the plus. Does anyone know where you can get a centronics to floppy convertor?
For all the latest Starquake remake news check out my website - www.endangermice.co.uk

Bryce

You just need to make the cable shown here: DIY:Floppy Drives - CPCWiki - The Ultimate Amstrad CPC Community & Encyclopedia!
Then you can connect the floppy to any of your CPCs and even a Spectrum +3 if you have one.

Bryce.

endangermice

That's exactly what I was after thanks Bryce!
For all the latest Starquake remake news check out my website - www.endangermice.co.uk

TFM

@Spirantho: Well, it's a bit sad that you don't have discovered the secret 360 rpm 3" floppy. However, it's a cool piece of hardware, since you can save data in an absolute secure way. As long as nobody physically takes your 360 rpm floppy, nobody else will be able to read your discs. Oh, how I would have needed such a device in the old days... ;)
TFM of FutureSoft
Also visit the CPC and Plus users favorite OS: FutureOS - The Revolution on CPC6128 and 6128Plus

Powered by SMFPacks Menu Editor Mod