Reviving a dead 1541

[srowe]

I've just dug my VIC-20 out of a long slumber. The computer itself is unscathed but the disk drive is non-functioning. On power on both green and red LEDs come on along with the drive motor. All three then stay on.

Looking around on the 'net and found some suggestions for tracking down the fault and this is what I know

• * The voltage regulator is putting out 4.93V, I think that's probably within the expected range.
  * The 6502 is good, I dropped it in the VIC and that booted fine.
  * I've swapped both VIAs with each other.
  * I've reseated the ROMs.
  * a visual inspection shows no obvious damaged components.
  * no IC is hot when power is applied

I'm wondering what the next step should be. I don't have spares for chips like the VIAs and ROMs and they're not easy to source.

[Spirantho]

I would guess the ROM is bad if you've tried the other bits, but it could also be the VIA.

I''ve just opened a spares section of my shop (see my post in this forum) for exactly this kind of thing. I should be getting in some 6522 VIAs soon, and I can already do a replacement ROM.

Where in the world are you? If you're in the UK you can send me your VIAs and ROM and I'll test them all for free. If one of them comes out faulty I can then supply the replacement, so you don't have to keep buying replacements until you find the culprit.

Just let me know if I can help!

[srowe]

If the ROMs are the most likely then I think I'll try and test them. I've got the bits to make a prototyping board, this will be my excuse to use them.

I've managed to snag a couple of VIAs on eBay. I'll also try and test these, any suggestions on suitable code to do so?

Thanks for the offer to test the chips. I am in the UK so if all my attempts fail I'll get back in touch.

[Spirantho]

MOS ROMs are quite notorious for being a bit unreliable. I've certainly seen enough fail over the years (much more than NEC/Hitachi ones for instance). If you can get a good dump from them, though, they should be OK.

Did you check the drive belt, by the way?

[srowe]

I'll probably build a ROM to EPROM adapter, there's ROM images on the 'net.

Not checked the belt yet but it's probably perished (the one in the C2N had). Any suggestions of where to get a replacement?

[Spirantho]

You can only get them from America, I think.

Having said that, I think I have one spare. Let me look tomorrow and I'll put one up on my new Spares section of my shop.

I got them from:
http://console5.com/store/parts/belts/c ... -belt.html

It cost me $17.85 for three, so about $6 each...
If you need one I'll do it for a fiver (it'll fit in an envelope so postage will be cheap) - but check to see if you need one first.
I'm trying to remember what the symptoms of a destroyed belt are - I have a feeling it may actually be what you describe (it's powering the motor but not sensing motor movement) but I could be quite wrong.
That said, I do have a 1570 that's not working - I need to check the belt in that yet.

Anyway, let me know if you need one and I'll see what I can do. If I don't have any left, I could do with some more so I could order some from console5.com so you wouldn't need to pay postage.

[srowe]

Thanks for the tip, I'd better check that before I get the soldering iron out.

[Spirantho]

I've been looking at my 1570 a bit, trying to diagnose its fault (motor running continuously).

The 74LS14 chip appears to be faulty, and apparently this is known as the "fuse chip" as it's likely to blow when the 1541 is disconnected from a C64 when it's running.

I'd check that chip before any others. Of course in my case the 74LS14 was the last of the 74LS chips on the board that I removed for testing... typical!

There's a thread about it here:
http://www.lemon64.com/forum/viewtopic.php?t=22529

I'm now waiting for some 74LS14 chips in the post - if you need one let me know. I do have a spare drive belt, by the way, should you need it.

Hope that helps!

[srowe]

I have a hunch my issue is the ROMs. From memory when I last powered on the drive >10 years ago it failed differently. I think it was the LED sequence that indicated the first ROM was corrupt.

I've nearly finished soldering my expansion port breakout board. Once that's done I'll try and dump the contents of them.

I took a quick look at the drive belt, it seemed ok. It certainly wasn't stretched and out of shape like the tape belt was.

[srowe]

Well an initial check of the top ROM seems to show it contains sensible values. Dumping it is awkward, I'd forgotten how primitive CBM Basic was. Even printing hex requires a page of code, and CRC computation is out of the question as we don't have XOR.

I think I might have to dump it to the serial port.

[Spirantho]

Without a checksum it's going to be very difficult to be sure of the ROM - if one of the address lines is flakey it'll look perfect to the human eye. You need either a checksum or a comparison it's going to be impossible.

You should just be able to add all the bytes together to generate a checksum, and then compare with a known good ROM's checksum...? Primitive but very unlikely to give a false reading.

[Mike]

[...] I'd forgotten how primitive CBM Basic was. Even printing hex requires a page of code, [...]

WTF? You can do it in two lines of code (see lines 3 and 4 in the example below):

Code: Select all

1 FORB=0TO255:GOSUB3:PRINTB,H$:NEXT:END
2 :
3 H$="":P=INT(B/16):GOSUB4:P=B-16*P
4 H$=H$+CHR$(48+P-7*(P>9)):RETURN

[...] and CRC computation is out of the question as we don't have XOR.

A XOR B can be expressed as either A AND NOT B OR B AND NOT A, or (A OR B) AND NOT (A AND B) in BASIC V2.

These two are really non-existent problems when it comes to checking the floppy ROMs. Actually reading out the memory of the drive with 'M-R' over the command channel - that thing is not entirely trivial, but nothing which couldn't be found out by taking a judicious look into the drive manual.

And regarding this:

On power on both green and red LEDs come on along with the drive motor. All three then stay on.

the FAQ in comp.sys.cbm states:

11.1. What do I do for my ill disk drive?

If the drive spins [...] [and i]f both LEDS are on:

Drive did not make it through the power-on reset sequence. Kernal ROM
(901229-05) is sus[pe]ct, as is (in order of expectation): VIA (6522),
CPU (6502), RAM (6116 or equivalent), or a "glue" chip in the reset logic

If your drive won't even accept input from the computer, and the drive
light is making some blinking pattern, then the drive may be telling you
what is wrong.

Did you check the RAM or the mentioned "glue" chip?

[srowe]

A small delay while I completed the RS-232 adaptor is started 30 years ago but the ROMs check out fine. I've also swapped out the 6522s with another pair, same result.

That leaves the RAM and other discreet logic that's soldered to the board. Ray's guide suggests UA1 and UD2 as candidates. Any other suggestions to try?

[Spirantho]

Did you check the 74ls14 yet?

[srowe]

That's UA1, I'm desoldering it now.