Level shifting circuits still not performing to my satisfaction. The reference driver logic from Diablo is a pair of MC858 NAND gates in tandem to sink 95 ma to pull a signal down to near 0V. The various gates I have tried don't have the drive power and few of the usual logic chips do.
I am going to take my time and try out various circuits until I have one that clearly produces the right results as measured by scope and voltmeter. I looked to see what circuits were used by other makers of drive interfaces for the same generation of drives.
DEC used a high current driver chip that could sink up to 300ma and I will try a similar chip, the SN75452 dual NAND driver which seems to provide similar or better performance. Fortunately I found I had 10 of them in stock, allowing me to breadboard one chip and do tests right away.
I am going to prototype and try each potential circuit on the real disk drive to ensure that my fpga, level shifter, cable and Diablo drive will produce the pulse voltages and shapes that I need. First up was the 75452B chip, which is an inverter thus requiring minor changes to the fpga for the two signal lines it will control.
With the logic inverted, I put on the scope and voltmeter, brought up the testbed and triggered a seek with those two cylinder values logically on. This will pulse them from ground to high when the seek machine fires, giving me a drop of the output from is high level to near ground.
It worked spectacularly, pulling the disk interface line down to near ground sharply. Time to rewire the board to use these chips for all the output lines and remove the level shifter circuits that were there. This will take a bit of time but I should be ready for real testing by tomorrow morning latest.
Having the potential that a max of 10 of the 13 lines could be swinging low at any time, I know I will be requiring about 1A of power supply to the driver board. I had to beef up the power supply wiring to handle the current.
No comments:
Post a Comment