Building the KLyball D-13

[BillO]

Thanks Dave. I thought it looked right too, but I am not sure. I couldn't even find an example of an FD data write signal on the web.
Given that it is correct, I'm still at a loss as to why an INIT would write only $24 all over a disk. I even tried another drive.

I'll see if I can dig up another cable to see if that might be the problem, but I'm seeing the right voltages when I probe the cable too, so I'm not confident that will help.

I should also put up the code I use and the process.

At the D/C/W/M prompt I choose C then load up the following code which initializes the floppy interface and writes $AA repeatedly to the ACIA.

Code: Select all

10 FD = 49152 
20 POKE FD+1,0
30 POKE FD,0
40 POKE FD+1,4
50 POKE FD+3,0
60 POKE FD+2,255
70 POKE FD+3,4
80 POKE FD+2,255
90 POKE FD+2,127
100 POKE FD+16,3
110 POKE FD+16,88
120 POKE FD+17,170
130 GOTO 120

On the scope, CH1 is connected to Pin-9 of J3.
Ch2 is connected to pin-8 of U69 and is chosen as the source for the scope sync.

[BillO]

I tried it out... just hangs... so Im looking at the code now.

/Jeff

Did you find anything with the code Jeff?

I'm still getting mysterious $24 written all over my disks, or so DOSDUMP tells me. I even verified the correct data patterns are being sent the the drive for each sector initialized ($43 $57 $TT $58) but all I read back is $24.

To make a long story short, out of shear desperation thinking the DOSDUMP program was running into memory issues I tried the memory test program. It fails at $3000 - says it's reading a $70 where is should be reading a $30 (bit 7 high when it should be low), but the program hangs there and does not return to the command mode so further testing or exercising the affected area cannot be done.

If you haven't found anything yet, I may look into it myself to see what may be amiss. It may just be memory problems causing it to fail too. Who knows?

[Klyball]

So the signal coming from 610 pin 9 going to floppy 22 is fine?
it works fine with the 600b but not 600d
how are you booting the system to get to A*
Do You get proper read with a disk already made?
have you probed all control signals coming from the 6821
What type of drives are you using?
are you using your memory board + 610 memory and or super board memory
Sorry for all the questions, Just trying to get a idea of the issue

I have written to successfully with 610 and d-13 -
Shugart 851 8"
Mpi-51 5.25
Toshiba-720k 3.5
Sony-1.44m 3.5
also 8" with 3.5 inch 1.44 as a,b,c,d

[BillO]

So the signal coming from 610 pin 9 going to floppy 22 is fine?

Yes, perfect!

it works fine with the 600b but not 600d

Well, it was not working 100% with the 600B because of the unique memory error that board is causing. However, I was able to initialize a disk correctly. I will put it back and see if that combination still works as before. That way I can make an initialized disk I can try to read with the 600D combination.

how are you booting the system to get to A*

On initial power-up I do a cold start to Basic, then load and run this basic program to initialize the FD system:

Code: Select all

10 FD = 49152 
20 POKE FD+1,0
30 POKE FD,0
40 POKE FD+1,4
50 POKE FD+3,0
60 POKE FD+2,255
70 POKE FD+3,4
80 POKE FD+2,255
90 POKE FD+2,127
100 POKE FD+16,3
110 POKE FD+16,88

Which are basically the steps the monitor takes before loading track 0 off a disk. Then I reset, go the the monitor and load up the following file:

BT.zip

(3.24 KiB) Downloaded 1908 times

Do You get proper read with a disk already made?

As yet unknown. I only realized the problem after re-initializing the disk I had been using before with the 600B combination.

have you probed all control signals coming from the 6821

Yes, no problem found there that I can see.

What type of drives are you using?

They are standard 5.25" SS/SD drives. They were sold as Radio Shack 26-3029 (and other similar numbers) as complete units with case and power supply. The drive in them are Texas Peripherals 10-5355-001 units and were used with Radio Shack Color Computers, Model 1, and as external drives for Model 3 and model 4. I have used them with my RS color computers and with old PCs too, without issue.

are you using your memory board + 610 memory and or super board memory

I'm using the 600D and 610 only.

Sorry for all the questions, Just trying to get a idea of the issue

Same here, and I have no issues whatever with you asking questions. I think maybe I'm asking too many though.

I have written to successfully with 610 and d-13 -
Shugart 851 8"
Mpi-51 5.25
Toshiba-720k 3.5
Sony-1.44m 3.5
also 8" with 3.5 inch 1.44 as a,b,c,d

I am sure it's something amiss with the build of one of my boards or one of the components I've used as both you and Jeff have had more success than I. This is also pointed to by the results I have from running that memory test, since swapping memory chips around the $3000 boundary did not make any difference to the test. This might indicate a build error or a non-memory component. In fact, from the outset of this project, the only problem I have found that was not a build error or component was an open trace on the A5 signal just under the socket at BASIC ROM1 on the 600D board. This looked like a run of the mill manufacturing flaw and was easily detected and fixed.

[Jeff]

Just an update on my Superboard-610-Klyball D13 boot issues.

I tracked the problem down to a faulty RAM chip in the 8K of RAM on the Superboard. I guess one (or more) of the 2114's just got too old and started to give up.

I now have my funky Epson Combo drive hooked up and full access to both drives (the 3.5 as Drive B/D and the 5.25 as Drive A/C). The Epson 5.25 drive is compatible with the two drives I hooked up to my C4P, so now I can transfer data between the two systems, Sneaker-Net style!

/Jeff

[lowrybt1]

Grant,

D-13 is built, 600D is working, 610 is working. Almost ready to connect a 5.25 floppy drive. Two questions I'm hoping you can help with:

1. What's the recommended procedure for calibrating TR1 when using the D-13 with the 610 board. What pins should I hook the probes to and what waveform\timing should I aim for?

2. I'll be trying this out with a SA850 drive, building an OS65D 5.25 boot disk by loading the image sideways using a serial connection from a PC (and using the process and files you so generously provided grant). Since the drive is not an MPI-51 and will have a different seek time and such, will I have any difficulty getting the os65 boot disk to work?

Many thanks, Tom

[Jeff]

1. What's the recommended procedure for calibrating TR1 when using the D-13 with the 610 board. What pins should I hook the probes to and what waveform\timing should I aim for?

Hi Tom,

The procedure for calibrating TR1 is detailed here, however I believe the timing for an 8" drive is different:
http://osiweb.org/osiweb/osiforum/viewt ... =220#p1266

/Jeff

[Klyball]

Tom, You can follow the link Jeff supplied but also can just hook up the drive, power it all up and and calibrate to give a 6us pulse from the Q output of the 74121 (pin 1).

But a SA850 is a eight inch drive I believe ,you will need to change things on the 610 if this is the case


Need to move the tx rx clk signal from 125 to 250
8 inch Drives
txrxclk=250kz

TXCLK= 260nS
TXDATA=260nS
RXCLK=1uS
RXDATA=2.75uS
data separator 2.75uS


5 1/4 Drives
txrxclk=125khz

TXCLK= 400nS
TXDATA=400nS
RXCLK=1uS
RXDATA=6uS
data separator 6uS

[lowrybt1]

Thanks Grant. Actually, I'm working with a SA-460, so 5.25 drive. Will report back on the TR1 calibration soon. Tom

[lowrybt1]

I hit a snag. 600B+610+D13 all connected but 610 board shorts (the PSU clamps down) when I apply 5V to molex pin 14 at the connection to the D-13. Being a relative novice may be the issue here. I am drawing the 5V directly from the 5V rail of the 610 board: soldered a wire to the rail and connected it (soldered) to the base of pin 14. Recommendations?