Hi! I'm Humberto from Argentina, volunteer at Espacio Tec retro computing museum. I'm currently working on getting a TRS-80 Model 4 up and running. The computer works OK, and I can load OS disks previously wrote using a PC+ImageDisk v1.19 and a known working Double Density YE-Data drive.
The Model 4 in question came with 2 identical (aside from one containing the resistor pack and not the other one) Shugart SA-400L drives. Both of them show identical behavior. I will from now on refer to them as "one drive" since both show equal symptoms:
- They won't read Model 4 disks, displaying "Disk Error" on screen.
- They will read Model 3 older disks (like one containing 13 Ghosts), but it will enter a boot loop after the AUTO option engage (I suspect a RAM problem but that would be another story).
They showing some life with Model 3 disks rings a bell about something having to do with modulation, I think.
Of course, if I swap the drive 0 with a known working YE-Data drive, it will load TRS-DOS and LS-DOS no problem.
Now the most important thing: IT WRITES DISKS CORRECTLY. Connecting the drive to a PC and using ImageDisk to write images, I could successfully make disks for TRS-DOS/LS-DOS that work on the YE-DATA drive and boot ok on the Model 4 using the YE-DATA drive. However, the Shugart won't read its own written disks. Also, using the align function in ImageDisk show the drive as "on track" for all 0-39 tracks, both on self-made disks and in disks written with the working drive.
Using the drive in the PC and a Model 4 emulator started from a virtual image, I can format a disk under TRS-DOS, but it will fail at writing the directory data (as it tries to read, it hangs up).
Things ruled out:
- Drive is lubricated and clean. Mechanism is in good condition.
- Head was cleaned with a cotton swab and isopropyl alcohol.
- Drive velocity was calibrated using ImageDisk RPM test and a disk written by another working drive.
- Track 0 sensor is properly aligned. It engages at track 0 position and disengages at track 1+ positions.
- Head alignment was checked with an oscilloscope on TP1 and TP2 signal points and dialed for maximum amplitude rotating the stepper motor.
- Resistor pack is installed and pull-up voltage is present at all the proper lines when drive activated.
The second most important thing is DATA IS PRESENT AT PIN 30 of the connector when reading attempt occurs. As long as I can see on the oscilloscope, it pretty much looks similar to the data coming out of the working YE-Data drive. So currently my most solid theory is that Read Data is there, but its quality is not good enough for the FDC to accept it.
Note that the Shugart drives won't boot both on the TRS-80 model 4 and they also will show similar "Disk Error" message when using them at the PC + Model 4 emulator.
I changed the only electrolytic cap C7 at the read circuit (check attached diagram for reference) with no luck. Removed cap showed no DC conductivity and its capacitance was spot on.
Any help and ideas will be greatly appreciated. I'm really stuck with this one.
Thanks!
Humberto
The Model 4 in question came with 2 identical (aside from one containing the resistor pack and not the other one) Shugart SA-400L drives. Both of them show identical behavior. I will from now on refer to them as "one drive" since both show equal symptoms:
- They won't read Model 4 disks, displaying "Disk Error" on screen.
- They will read Model 3 older disks (like one containing 13 Ghosts), but it will enter a boot loop after the AUTO option engage (I suspect a RAM problem but that would be another story).
They showing some life with Model 3 disks rings a bell about something having to do with modulation, I think.
Of course, if I swap the drive 0 with a known working YE-Data drive, it will load TRS-DOS and LS-DOS no problem.
Now the most important thing: IT WRITES DISKS CORRECTLY. Connecting the drive to a PC and using ImageDisk to write images, I could successfully make disks for TRS-DOS/LS-DOS that work on the YE-DATA drive and boot ok on the Model 4 using the YE-DATA drive. However, the Shugart won't read its own written disks. Also, using the align function in ImageDisk show the drive as "on track" for all 0-39 tracks, both on self-made disks and in disks written with the working drive.
Using the drive in the PC and a Model 4 emulator started from a virtual image, I can format a disk under TRS-DOS, but it will fail at writing the directory data (as it tries to read, it hangs up).
Things ruled out:
- Drive is lubricated and clean. Mechanism is in good condition.
- Head was cleaned with a cotton swab and isopropyl alcohol.
- Drive velocity was calibrated using ImageDisk RPM test and a disk written by another working drive.
- Track 0 sensor is properly aligned. It engages at track 0 position and disengages at track 1+ positions.
- Head alignment was checked with an oscilloscope on TP1 and TP2 signal points and dialed for maximum amplitude rotating the stepper motor.
- Resistor pack is installed and pull-up voltage is present at all the proper lines when drive activated.
The second most important thing is DATA IS PRESENT AT PIN 30 of the connector when reading attempt occurs. As long as I can see on the oscilloscope, it pretty much looks similar to the data coming out of the working YE-Data drive. So currently my most solid theory is that Read Data is there, but its quality is not good enough for the FDC to accept it.
Note that the Shugart drives won't boot both on the TRS-80 model 4 and they also will show similar "Disk Error" message when using them at the PC + Model 4 emulator.
I changed the only electrolytic cap C7 at the read circuit (check attached diagram for reference) with no luck. Removed cap showed no DC conductivity and its capacitance was spot on.
Any help and ideas will be greatly appreciated. I'm really stuck with this one.
Thanks!
Humberto
