• Please review our updated Terms and Rules here

IEEE488 and 1541 drive, the IEEE interface...

giobbi

Veteran Member
Joined
Dec 23, 2012
Messages
987
Location
São Paulo country, Brazil
Hi,

Reading an old post and the web page http://vintagemashups.net/2010/12/commodore-1541-ieee-488-conversion/ about how to convert a 1541 drive and make it suitable for PET, using the IEEE-488 protocol, I decided to do a try. So I bought the needed ICs, I built a pcb and, as usual, I have a little, simple problem... nothing works :smile:: no answer calling the drive from my 8032-sk and no reset when I switch on and off the 8032.

Of course I checked my PCB (and I will check it again and again), but I've noticed the drive has a strange behavior with the patched ROM image I downloaded from the above site.

- with the original ROM, as usual, the floppy starts to spin and the red light comes on; after few seconds, the light and the spin motor go off in the same moment. Normal behavior.

- with the patched ROM, the floppy just spins for a while and then stops, while the red light stays on for a longer time (and then goes off).

Note that the strange behavior doesn't depend from the interface board; I tried to put the patched EPROM and leave the 6522 into its socket, without the ieee interface board, and the behavior is the same. I think it's quite strange: Shouldn't it have the same reset procedure, even with the modified ROM image?

I'm only sure that:
- the 1541 drive is ok
- the 6522 is ok
- the socket adapter is ok
- the EPROM im programmed returns no difference if compared with the patched rom image (so I believe it's ok too)
- the ieee cable an connectors are ok

then, probably, the issue is with my board or with the patched ROM image.


So here there are my two question for you, smart people:


1) Anybody that understand ML, can take a look to the patched ROM from the vintagemashup.net site, compare it with the original ROM 901229-05 and tell me if this different behavior during the drive boot is ok? I know it's a lot of work, but maybe somebody here likes to compare hex files? ;-)

2) looking at schematics and at the project (http://www.zimmers.net/anonftp/pub/cbm/documents/projects/ieee-488/index.html), can you tell me how the jumpers work? i.e. no jumpers = drive 8, all jumpers = drive 11, etc.?


thank you,
Giovi


p.s. OH OH OH I have a scope! Finally, I have a scope! I'm not sure if it would be useful for this problem, but I have it :)
 
Hi,
...
2) looking at schematics and at the project (http://www.zimmers.net/anonftp/pub/cbm/documents/projects/ieee-488/index.html), can you tell me how the jumpers work? i.e. no jumpers = drive 8, all jumpers = drive 11, etc.?
Haven't looked at that in detail, but it's usually the opposite, i.e. both jumpers=8 and cutting a jumper adds its binary value (1 or 2).
p.s. OH OH OH I have a scope! Finally, I have a scope! I'm not sure if it would be useful for this problem, but I have it :)
OH OH is right; now we can really look forward to some interesting 'projects'... ;-)

Sounds like you're having a lot of fun!
 
Haven't looked at that in detail, but it's usually the opposite, i.e. both jumpers=8 and cutting a jumper adds its binary value (1 or 2).

ok, I will try with both jumpers. Unfortunately often people that develops this kind of project hasn't patience enough to care of dummies with complete info and instructions, LOL

OH OH is right; now we can really look forward to some interesting 'projects'... ;-)

ok, next week I will have a trip to Italy, and hopely I will be back at home in november with a PET 2001, an 8032 and two drives (8050 and 8250). Of course none working, so keep ready :)

Sounds like you're having a lot of fun!

Sounds like I'm having a lot of problems, as usual ;-) , but you're right, it's funny :)
 
Well, I don't know why it doesn't work and I don't know what to do more..... :wallbang:

From the german magazine:
Here there is the original schematics: http://www.verrua.org/1541-ieee/magazine-schematics.jpg
Here the part list: http://www.verrua.org/1541-ieee/magazine-list.jpg
here the original pcb: http://www.verrua.org/1541-ieee/magazine-pcb.jpg

And here a better schematics: http://www.verrua.org/1541-ieee/schematics.jpg
(note that the schematics has the pins 12 and 13 of the 76136 swapped respect the magazine, but it shouldn't have any influence).

Since I found the original PCB too hard for a DIY pcb made using iron and laser printer, I designed an easier pcb layout with Eagle: http://www.verrua.org/1541-ieee/easy-pcb.brd

and I used this image: http://www.verrua.org/1541-ieee/ieee488rom.bin


A friend of mine, Ruud Baltissen, told me he did that project in the past too, and it's working fine (He sent me the magazine pdf and its version of the eprom image).

I bought the 75160 and 75161 ICs on eBay. I hope (and believe) they are ok; I bought a couple per kind and I believe at least one of them should be ok. I also tried two different 74136 and 6522.

I don't know how many times I checked what I've done; I checked everything without to find any mistake. I've done the PCB twice, just to be sure. I also changed the ieee connector.

The 1541 works fine with its original rom and 6522; the ieee port on the 8032 is working fine too.

I tried many commands like: load"$",8 - directory d0 -directory d0 onu8 and the format and initialize commands used from C64.

If I try the command the first time, I only get a ?device not present error; if I try the same command again, the computer hangs until I don't switch off the drive; then I get the device not present error.

Honestly I don't know where I went wrong..... :-(

Any idea?
 
Well, I don't know why it doesn't work and I don't know what to do more..... :wallbang:

From the german magazine:
Here there is the original schematics: http://www.verrua.org/1541-ieee/magazine-schematics.jpg
Here the part list: http://www.verrua.org/1541-ieee/magazine-list.jpg
here the original pcb: http://www.verrua.org/1541-ieee/magazine-pcb.jpg

And here a better schematics: http://www.verrua.org/1541-ieee/schematics.jpg
(note that the schematics has the pins 12 and 13 of the 76136 swapped respect the magazine, but it shouldn't have any influence).

Since I found the original PCB too hard for a DIY pcb made using iron and laser printer, I designed an easier pcb layout with Eagle: http://www.verrua.org/1541-ieee/easy-pcb.brd

and I used this image: http://www.verrua.org/1541-ieee/ieee488rom.bin


A friend of mine, Ruud Baltissen, told me he did that project in the past too, and it's working fine (He sent me the magazine pdf and its version of the eprom image).

I bought the 75160 and 75161 ICs on eBay. I hope (and believe) they are ok; I bought a couple per kind and I believe at least one of them should be ok. I also tried two different 74136 and 6522.

I don't know how many times I checked what I've done; I checked everything without to find any mistake. I've done the PCB twice, just to be sure. I also changed the ieee connector.

The 1541 works fine with its original rom and 6522; the ieee port on the 8032 is working fine too.

I tried many commands like: load"$",8 - directory d0 -directory d0 onu8 and the format and initialize commands used from C64.

If I try the command the first time, I only get a ?device not present error; if I try the same command again, the computer hangs until I don't switch off the drive; then I get the device not present error.

Honestly I don't know where I went wrong..... :-(

Any idea?
Sounds like a worth while project, considering how cheap and how many 1541s are out there compared to 2031s or even 2040s, 4040s or 8x50s. I may try it myself but I don't really need one; why doesn't another PET owner give it a shot?
 
Sounds like a worth while project, considering how cheap and how many 1541s are out there compared to 2031s or even 2040s, 4040s or 8x50s. I may try it myself but I don't really need one; why doesn't another PET owner give it a shot?

Don't you love challenges? Try it by yourself (using my pcb layout of course) and discover where I went wrong! And then tell me, so I can fix it ;-) LOL
 
Ok, I give up.

I can't make it to work, I believe I tried everything that passed in my mind. The only thing I can imagine is the ICs I bought are wrong or broken, but honestly I think it's quite hard to believe (and I bought two per kind). The 8032 ieee port works fine, the 1541 works fine, the eprom should be ok (I dumped and patched mine using the original basic program from the original .d64 magazine disk), I checked the pcb I designed so many times.... booooh.....

I'm sure the project work, many clever people did it, so of course the fault is mine.

Please, let me know if somebody will achieve some result with it....

--Giovi
 
Sorry to dig up an old thread, but has anyone got this to work?

I have also done the IEEE-488 conversion on a 1541 and mine isn't working either. It is behaving as in the thread, the power on disk seek is shorter than normal and nothing is read on the Pet. It seems to be working in that the activity light on the drive goes solid when you run a command to ID 8, then starts flashing. Sending a command to ID 9 doesn't affect it.
1541 mods.jpg
Trying to connect via an xum1541 type USB interface, I do get a response, '73,iec-dos v2.6 2031,00,00' even occasionally a full directory listing. But it only works once then stops. (the interface works fine on an 8250)

As with the original poster, I've checked the 1541 (two 1541's in fact), the ROM, my wiring etc.

I've also tried using an image of the original 2031 ROM set, but the disk just spins and doesn't stop. I didn't really expect that to work as there are some differences between the 2031 circuit and the 1541 with the plug in board. I presume these were accounted for in the original patched ROM.

Thanks,

Dave

More details on my blog here:
http://blog.tynemouthsoftware.co.uk/2015/01/commodore-ieee-488-disk-drives.html
 
I have actually gotten it to work. It might even have been me who drew the new schematics.
Maybe the trick is to use the provided patch for the 1541 ROM, and not a real 2031 ROM. You can download the schematics as well as the patch on the link here http://www.6502.org/users/andre/cbmhw/ieee488/index.html

OK, thanks, good to know it should actually work. Now I need to work out what bit I got wrong.

The 2031 ROMs were worth a try after I thought I'd ruled out everything else. I'm using the original 325302-01 ROM and the ROM image I got from vintagemashups.net/2010/12/commodore-1541-ieee-488-conversion (the page seems to be down now). This image also matches the one linked earlier in the thread on verrua.org.

Unless I've missed it, the archive on 6502.org doesn't contain a ROM image, only the program to patch the 1541 ROM. The appears to be the same as the one I got from zimmers.net (http://www.zimmers.net/anonftp/pub/cbm/documents/projects/ieee-488/index.html).

Thanks,

Dave
 
I have actually gotten it to work. It might even have been me who drew the new schematics.
Maybe the trick is to use the provided patch for the 1541 ROM, and not a real 2031 ROM. You can download the schematics as well as the patch on the link here http://www.6502.org/users/andre/cbmhw/ieee488/index.html

Thanks, it's good to know it does work. So it's just I've got something wrong. Although I still can't see any problems with my wiring, I've been through it several times, made up a new PET-GPIB lead and tried swapping out all the chips. Trying out the full 2031 ROMs was worth a try, but wasn't expected to work.

Unless I've missed it, there isn't a ROM image in that archive, just the BASIC code to patch the ROM from a C64. I've been using the ROM image from the vintagemashups post (which seems to be down now), and that matches the one on verrua.org linked in an earlier post. I've also checked my wiring against those schematics.

I'm running out of ideas. The one that is left is to try to find a good clear 1541 schematic for one of the drives I have spare (1540050 or 251830) and see how much difference there is and try to fully convert it to match the 2031LP.

(I did reply on Sunday but that doesn't seem to have appeared yet, so I'm repeating myself, but without the links in case that was the issue).

Thanks,

Dave
 
Just a stupid question, looking at the picture of your conversion. Did you use the correct 6522?
From my memory (my own 1541 is stored away) I think I remember that my conversion board was set more to the side, under the second 6522 from the CPU. I may be wrong, but maybe you wanna check.

Also: IIRC on some newer drives there was a track-0 sensor that used one of the parallel port bits used by the IEEE488 conversion. So you may want to check that too.
But the 1540050 should be fine, as it's old enough...

André
 
Just a stupid question, looking at the picture of your conversion. Did you use the correct 6522?
From my memory (my own 1541 is stored away) I think I remember that my conversion board was set more to the side, under the second 6522 from the CPU. I may be wrong, but maybe you wanna check.

At this stage these are the sort of things I'm rechecking. I'm fairly sure it's the right one as I have been able to get directory listings out of it, I doubt that would happen if it were in the wrong socket. I've just a long legged wire wrap socket for the 6522 and cut off the pins that aren't meant to pass through. This leaves it directly above the original socket rather than offset to the side if a separate socket and pins were used. Again, pretty sure it's the right set of pins.

The only one I wasn't sure about is PB5 (pin 15). On the 2031 this is head select (which is why the 2031 ROMs don't work here). On all the schematics for the 1541-IEEE I've seen and checking back to the PCB from the original magazine this pin is not connected.

I've also tried a selection of 6522's, 6502, driver chips etc. Checked all the wiring through again. With all that ruled out, it is probably something stupidly obvious that I've missed.

Thanks,

Dave
 
At this stage these are the sort of things I'm rechecking. I'm fairly sure it's the right one as I have been able to get directory listings out of it,

So what are your actual symptoms if you can get a directory listing from it? In a previous post you wrote you got a device not present error.
The device not present actually points to the XOR logic.
The only one I wasn't sure about is PB5 (pin 15). On the 2031 this is head select (which is why the 2031 ROMs don't work here). On all the schematics for the 1541-IEEE I've seen and checking back to the PCB from the original magazine this pin is not connected.
I don't think PB5 is connected on the 2031 either. The 2031 only has a single head.
I've also tried a selection of 6522's, 6502, driver chips etc. Checked all the wiring through again. With all that ruled out, it is probably something stupidly obvious that I've missed.

Did you check that the diodes are in the correct orientation?

Do you have a scope to check the signals? You can trigger on falling ATN and check that NRFD and NDAC go low immediately after that due to the XOR logic.

André
 
Thanks for the suggestions, and sorry for any confusion. I started with a number of IEEE-488 devices, none of which would talk to each other. Once I finally got one pair talking, I was able to identify the faults in the others. I now have an 8250 drive talking to an 8032, a 4032 and via a USB adapter. That seems to work consistently.

The 1541 converted to a 2031 works some of the time. It is intermittent, but it doesn't appear to be intermittent as in a loose wire sort of way, more memory fault or floating input type thing. The symptom is sometimes it works fine, then the next command fails, several more may fail then it works again. I see the same thing on a real Pet and using the USB adapter, so I am going to point at the drive.

Here are some screen shots of repeatedly asking for a directory. Sometimes it is not detected, and I need to reset or I get device not present errors. Sometimes it is detected, but the directory command returns nothing. Other times it returns the header and the number of blocks, but no files. Other times it will give a full directory listing. Usually in that order, but not always.
intermittent dir.jpg8032 intermittent.jpg

I've transferred the 2031 demo disk image and there is a drive performance test on there. I've run this a few time, sometimes it passes. Sometimes it fails and the drive locks up (flashing access light).
Drive test fail.jpgDrive test pass.jpg
I've tried this on different 1541 base drives, and with both the on board memory and ROMs and with those bypassed using a ROM/RAM board in the CPU socket. I get the same behaviour with all of those.

That is what lead me originally to suspect it may be an issue with the ROM image itself, but if others have it working reliably then maybe not.

I have checked the diodes again just to be sure, and they are 1n4148 diodes oriented according to all the schematics I have seen. I'll monitor the suggested pins and see if I can see any difference between success or failure.

Thanks,

Dave
 
Ah I see.

The symptoms of the directory stopping or not coming at all seems to point at the ATN response. What does the "ST" BASIC variable give after an interrupted directory?

You should know that the DIRECTORY command is about the most inefficient implementation of the command. It basically continously sends a "TALK", gets a byte, then "UNTALK" in a large loop... So there is a lot of activity on the ATN line - which formidably tests the XOR arrangement. As it seems to work some times, other times not I don't think it's the IC. How long are your cables? How many devices are on it? Can you measure the signals on the ATN/NRFD/NDAC lines with a scope? Is the ground connected well? It could be that if your cables are rather long or there is "unusual" load the Open-Collector outputs may not drive the NDAC/NRFD lines as low as it should be.
Note that the older dual drives use MC3446 drivers for IEEE488 (like the PET), only the 2031 uses the 75160/75161. The difference is that the XOR arrangement on the 75160/161 is directly on the bus lines, while the MC3446 arrangement has it on the TTL logic side, not directly on the bus.
If you have multiple drives etc (even if off) on the IEEE488 bus, can you remove them and only test with the PET and the 2031?
 
Thanks for the advice, getting there. Things have changed slightly now though, I can't get it to fail on the PET any more. It ran without a problem for a couple of hours, and I wasn't be able recreate the problem. The ST variable held 64 after successful directory, but I couldn't get it to fail on either PET. Whilst this is a good thing, I'd rather have changed something to make it work rather than it just starting to work by itself, but if it continues, I'll accept it.

I had been testing with a single short cable from computer to drive, everything else disconnected to rule it out. I made a shorter single cable last week to be sure, and it was still failing until last night.

I hadn't realised DIRECTORY was that bad, I was trying all sorts of things, just wanted to give a clear example of three commands in a row giving different responses. Also, I'm going between a 4032 with a N keyboard and an 8032 with a B keyboard, so anything with "$" in it gets me every time when I move between them as it doesn't require shift on the graphics keyboard, but does on the business one. DIRECTORY is the same on both.

However, it still fails on the USB adapter xum1541/zoomfloopy style thing, before and after the successful tests on the PETs. I've checked over the ATN wiring again on both sides and can't see anything amiss.

I've monitored the ATN, NDAC and NRFD lines and some others. This shows the states during a reset and then 5 dir commands. The first two returned full listings. The third and fifth returned nothing, and the fourth returned a partial listing. This was all within the space of a minute. There are a few more detailed sections of these, but they seem to show that ATN doesn't stay low on the two fails.

Reset, OK, OK, OK, nothing, patial, nothing.jpgpartial.jpgno response.jpgok.jpg

Thanks,

Dave
 
Just a small addendum to that, I now have seen the same error again on the Pet. The ST variable holds 64 after success, and 2 after a fail. After these initial failures, it then worked fine for half an hour.
At power on.jpgST=2.jpg
Thanks,

Dave
 
Back
Top