OSI floppy disk creation with autoDC

[BillO]

Not much to report..

I can get the disk tool loaded up fine and can select between XMODEM and XMODEM(CRC) and have tried both, but cannot get the transfer to occur or even start.

When I get to the point where it asks me to press any key to start the transfer, I press a key and follow up immediately with invoking the Tera-Term XMODEM transfer, but nothing happens. The transfer just times out and by then the disk tool has become unresponsive and will not respond to ctrl-X or esc. In fact, the disk tool becomes unresponsive right after I press a key to proceed.

I tired the similarly designed disk testing tool. I can also get that to load an run too and it seems to write to the disk fine (at least goes through the
motions), but once it begins to read back it crashes the machine hard! Even wiping itself out of memory.

[Mark]

I'm sorry to hear you are having problems.
I'm using this version of TeraTerm

Teraterm version

teraterm.gif (12.95 KiB) Viewed 16637 times

Although OSI defaults to 8N2, I forgot that DiskTool uses 8N1 communication parameters (not that it should matter).
I'm using these parameters.

Comm settings

teraterm2.gif (14.28 KiB) Viewed 16637 times

Generally, I load DiskTool with TeraTerm "Send File..." disktool.lod, put a floppy in and then Transfer/Xmodem/Send, select my disk 90K OSI image & go.

How are you interfacing to the host computer? Are you using a USB-Serial adapter? If so, what brand? [Some FTDI brands are clones which can be detected by the official drivers and are intentionally corrupted. ] Not that this would cause the crashing problems you've seen with DiskTest.

If you have time, please try David's official version here. (Recent changes could have introduced a bug.) Does RPM test work?

What does the OSI display say when you run DiskTool? What is MemTop?
Can you try DiskTool again, select option #7 to use HEX transfer and try #1 Dump Raw Image?

Thanks!

[BillO]

Hi Mark,

Thanks for the support!

I'm using the same process as you describe except I'm at 4800 baud. Some day I'll have to speed that up.

My version Tera Term 4.105. I guess I could look for an older version.

I'm using COM1 which is implemented in a integrated UART on the MB.

I tried both 1 and 2 stops bits, but it made not difference. Suspecting the 610 board I tried using my own memory expansion just to see if I could kick off the transfer, but it behaved exactly the same way. I also tried both SYNMON and CEGMON - no difference in behavior. I even tried swapping out the 6502 and the 6850 on the 600 board. Exactly the same thing. This one is a 600 rev. D, if that makes a difference?

It's normally a pretty rock solid machine and I can usually do all sorts of I/O work on my BUS expansion running up to 6 devices at a time.

I did not try the disk speed test - I'll do that tomorrow. I'll also try Dave's version.

I'll also try the dump raw image.

Historically I have had no luck with getting the 610 and a drive or two going. I also have a 600 rev. B, but it won't work with that either. That one has developed a strange issue now. I was going to post about that soon. I'll update on the disk issue tomorrow.

Thanks again Mark.

[BillO]

So, update.

For DiskTool:
MEMTOP is $7FFF
Selecting 7 > H, then 1 > Return, the machines again just locks up.

For DiskTest (official version):
The RPM test sort of works. It returns 122(+) for 5.25 drive and 244(+) for a 3.5" drive. However I know those values are wrong as both these drives are everyday working drives. Maybe swap out the 6850 on the 610 board?

I have to run an errand, but will try that later.

[Mark]

Those RPM rates seem to indicate a hardware problem of some sort.

The floppytest RPM test works by watching for the falling edge of the disk index signal, writing data to the disk ACIA and counting the number of bytes written at disk I/O speeds until the return of the index signal (16 times), then calculating the RPM for display.

The disktool RAW dump code reads data from the disk ACIA, starting at the end of the index signal until the index signal returns. On 5.25" this should be up to 2159* bytes or so. If there is no data being read at all by the disk ACIA, it should still abort at the next index signal. I'm not sure why it is hanging. It is not a console I/O problem since you have no trouble seeing menus & selecting options, so it's likely not getting to the point where there is data to send. Even a bulk erased disk should have noise on it to trigger a few bytes of data from the disk ACIA.

So maybe its a PIA or buffer problem with the index signal or disk ACIA problem.... ? I'll have to look closer at the code to figure out what is going on & why it sends nothing.

With the floppy disk tool, you can select #8 Status screen. You can load/unload head step up & down & monitor PIA signals. You should see the index signal flash between 0 & 1 occasionally as it catches the disk rotating while the disk is selected.

Good Luck!
-Mark


* The clock rate for 5.25" should be 125KHz (125000 bits/sec) the default disk ACIA word format is 8E1 1 start bit, 8 data, 1 parity, 1 stop bit or 11 bits total. A regular 5.25" drive should rotate at 300RPM or one rotation every 200ms. 125000*.2 sec is 25000 bits / 11 bits per serial word = 2272 bytes MAX. But there is a delay of 10ms or so after index before data is written by OSI so that leaves (125000*.19)/11 or 2159 bytes of space per track not including other delays between track header & sectors etc.

[BillO]

Thanks Mark,

In the mean time I'll try to swap out the disk ACIA. The PIA is the only one I have. I'll have to order in a "new" one off eBay.

[BillO]

I swapped the disk ACIA for the COM ACIA and got a bit further.

DiskTest:
The speed test now shows 305.25 RPM for the 5.25" drive which is a bit more reasonable. Too bad it's not a continuous readout to help with adjustment.

The disk test gets a tad further too. The machine no longer crashes completely on the first read, however the read returns:

RTRK 0 -2167

Then promptly locks up.

DiskTool:
No change in behavior from my last post. Still locks up whenever a transfer is attempted and it will not access the disk choosing option 1.


Interestingly, the ACIA I pulled out of the 610 board I used in the 600 board and it works just fine there. I noticed that the socket I used for ACIA on the 610 board was a really poor quality single wipe socket (the only one on the whole board) so I replaced it with a machined pin socket.

Now I need to try to swap out the PIA. It's in a good socket, but it may be a partial dud I guess. I don't have any handy but I have just ordered some.

[BillO]

In waiting for the new MC6821 to arrive I decided to give some time to my Rev B unit. It turned to be a simple repair. There was a crack in the A8 trace between the monitor socket and the ACIA.

I moved the 610 board over to that and loaded up the Disk Test program. On this machine it was able to read back the whole disk without locking up and gave the output:

RTRK 0 -q399
RTRK 1 -q399
RTRK 2 -q399
.
.
RTRK 39 -q399

I have no idea what -q399 means. The documentation is sparse and does not even mention (AFAICS) "q" at all.

I moved on to trying DiskTool. It loads up, but won't run. It just send a couple of garbage characters to the terminal and locks up. Does not even report RAMTOP on the OSI video display.