New issue cropping up? 1541 Mark ii power supply failure

[br44]

Hey guys, just thought I would pop in and let people know what I discovered recently.

So the story is... I recently won a 1541 mark II drive on ebay. Complete in box with all the paperwork, cables, introductory floppy, and original power supply. It seems the fella who sold it had cared about it enough to keep everything all these years.

In my excitement, I pulled the drive out and got ready to test it. Oh how I planned on lovingly restoring it, lubing the rails, and cleaning the head. You know... all the fun stuff people do for floppy drives. Well as it turns out, I'm a paranoid fella. I used a multimeter to check the voltages on that epoxy brick that came with the drive. These drives use two voltages, 5v DC and 12v DC. The 5v DC was happy, coming in at a nice 5.1 volts. That 12v rail though... was a very limp 7 volts. As soon as I connected it to the drive and hit the GO switch, that 7 volts immediately collapsed to ZERO. The drive did not spin, and both lights stayed illuminated on the front of the drive. Typical behavior for any 1541 is for the drive to spin briefly, both lights to illuminate, and one light goes out after a couple of seconds.

The next day was spent pulling apart the epoxy brick, only to find that it is indeed impossible to fix. There's no way to get at the undoubtedly bad 34 year old capacitors that were in the brick. So I was forced to salvage the wires and plastic case, and throw away the brick. Then I spent all of today looking in my stash of junk electronics for a suitable power supply for my 1541. In the end, I settled on an old cell phone charger and a wall wart for an old PC scanner. I verified both provided voltages in a safe range, and am now in the process of wiring them up and fitting them in the original plastic casing. I did test it, and the drive seems to function properly now.

I thought I would tell you all about this, because this seems to be a new issue coming up. Everyone and his brother knows that the bricks for the C64 are dangerous and like to fail (and destroy computers), but I've only ever seen a single person mention the 1541 Mark ii bricks having a problem. That man... and his video on the subject... are linked below.

 

[robert_sissco]

Yea, as a general rule I just never use any power supplies from Commodore. My VIC-20's, C64s, C128, Amiga and 1581 all have brand new modern power supplies. I have not heard of anything good from those old power supplies and I would just rather may a bit more for that piece of mind.

 

[Hugo Holden]

I did not know that some 1541 drives had an external psu. The two 1541 units I have have a very robust internal transformer and a well designed regulator system, also easy to repair if required. But of course they are nowhere near as compact as the MK II unit.

When I looked on ebay, it appears there is a seller, selling a version of the psu, which may well be better than the difficult to service original supplies:

Commodore oc118 Power Supply Adapter | eBay

Find many great new & used options and get the best deals for Commodore oc118 Power Supply Adapter at the best online prices at eBay! Free shipping for many products!

www.ebay.com

 

[br44]

I did not know that some 1541 drives had an external psu. The two 1541 units I have have a very robust internal transformer and a well designed regulator system, also easy to repair if required. But of course they are nowhere near as compact as the MK II unit.

When I looked on ebay, it appears there is a seller, selling a version of the psu, which may well be better than the difficult to service original supplies:

Commodore oc118 Power Supply Adapter | eBay

Find many great new & used options and get the best deals for Commodore oc118 Power Supply Adapter at the best online prices at eBay! Free shipping for many products!

www.ebay.com

The issue with the original 1541 drives and their internal power supplies is that they produce a lot of heat, and heat isn't good for IC chips. They tend to get roasted alive, especially if someone throws papers, floppies, another drive, their lunch.... on top of the vents of the 1541. The mark ii drives naturally don't have that problem because the power supply is external, plus they are made later in the life of commodore computers so the chips tend to be more reliable (MOS finally figured out their chip making process). That's why I went with it, as opposed to a normal 1541.

Not that I'm saying anything is wrong with the iconic 1541... it's just they have a higher rate of chip failure than the mark ii. Otherwise they are the same drive internally, with cosmetic differences.

 

[Hugo Holden]

Yes it is interesting when there are ventilation slots on the top of an instrument, especially if they are on a fairly broad horizontal surface and the unit relies on convection rather than Fan cooling. Sooner or later somebody comes along and uses the surface as a shelf for A4 paper. The IMB5153 VDU has this issue/possibility.

If that is not enough, upper slots on a horizontal surface are more prone to let dust into a unit. Often it is better to have the ventilation slots high up on the vertical sides of an instrument, this way they let in a lot less dust and rarely get occluded by accident.

I did not know that the 1541's with the internal power supply had thermal problems and a high chip failure rate as a result. I hope my SFD1001 doesn't suffer the same fate, maybe I should install a small Fan to increase the convection cooling.

One interesting thing about IC's and thermal related failures, it appears in some cases to be not just the magnitude of the rise above ambient temperature that causes delamination & other failures in the chip, but the number of times the chip is thermally cycled, it is the product of the two values that appears to be important.

So a chip can appear to have a reasonable rise above ambient temperature on thermal testing, which doesn't seem excessive, but if it is power cycled like that over a few decades then the chip fails. (There is a classic chip failure in the Tek 2465B scope in the H deflection IC U800, where one of the output transistor pairs loses its bonding). For some IC's, if the magnitude of the thermal cycle above ambient temperature can be reduced (with heat sinking, fans etc) the chip will have a much longer life. Obviously if the chip is not thermally cycled at all (in a storage drawer as a spare part) it would have an exceedingly long life.

One thing with my 2465B I have been taking screen shots of it over 15 years or more. Going back and looking at the old photos, from old articles I wrote, I can see the failure in evolution, with a slow, progressive left shift of the data display (a small progressive DC offset in the output), until finally one of the output transistors in the IC went open and the display jumped off the screen. It is really unusual to have a recording of a fault in evolution, over a decade of thermal cycling. I did not notice the display position changing on daily use with the scope.

 

[br44]

I did not know that the 1541's with the internal power supply had thermal problems and a high chip failure rate as a result. I hope my SFD1001 doesn't suffer the same fate, maybe I should install a small Fan to increase the convection cooling.

If I had an original 1541 handy, I would design a fan bracket that is 3d printable. I did that with my breadbin c64, where it mounts up in the top case with double sided sticky tape (so it doesn't permanently alter the case). it wouldn't solve the problem with covering the vents, but at least it would hurry up the air exchange inside the case.